It's interesting how anemic the use cases seem to be - we see the same things recycled over and over: "reword my email", "remove object from picture", "add a reminder", "summarise my text message which was already only 20 words long" etc etc. As if these are the major problems in people's lives.
I really feel like there's a fascinating valley of death between simple things that actually work and things of real value that are actually still beyond the horizon. They either aren't reliable enough, aren't accessible to the tech, or exceed the sophistication of our existing trust models. For example, I'm planning a trip. Booking a multiday holiday - there's a real beast that is time consuming, complex and painful. I test out the AI tools. They fail. Hard. Hallucinations all over the place, false confidence, inability to act, inability of me to trust their actions.
It's just nowhere near practical utility yet. Not "nearly there" but "not nearly half way there". I got the top tier of Gemini AI. Can it rent me a car? "As an AI I can absolutely guide you through the process of renting the car, but I can't physically access the web site or type in the details for you".
An alien landing on earth consuming Apple marketing content would be under the impression that humans did nothing but organise hikes to Big Sur with their friends.
And even if I organized a hike to Big Sur, Apple Maps would be pretty worthless because it doesn't have trail visualization and navigation worth speaking of (ok, maybe in Big Sur, but certainly not the rest of the world).
So I end up pulling out the trusty old Garmin gpsmap with cycle/hiking maps, that survived drops from 1.5 meters at 30 km/h as I was gliding of a mountain with my bike.
I recommend mapy.com (mobile app and web app too when on computer) - they mostly use OSM data and rendering of map tiles is great. Also offline maps etc.
There are several great options, besides mapy, e.g. Organic Maps and CoMaps also work pretty great. There are also some really good bike optimized apps (like NodeMapp for the Dutch cycle network).
But I generally prefer to use a Garmin GPS or watch. They work for days without charging (the older models even work with two AA batteries), very robust (e.g. their gpsr survives drops), work well offline, and transflective displays work better in direct sunlight.
For planned routes, I make then in NodeMapp or some other focused application and send the GPX overs to a gpsmap unit or Fenix watch. Many national parks, etc. also have great GPX files for recommended hiking/cycling routes.
I remember a couple of years ago it was all about booking a table at a restaurant, as if this would be the main spare time activity of people in the Apple bubble. I wonder when the shift from restaurant bookings to outdoor activities happened. Maybe a pre- vs post-Covid thing?
As Psychology is the study of the psyche of white male psychology students, big tech is technology for affluent citizens of San Francisco with a career in IT…
Even the aspirational use cases you're talking about basically are just "digital secretary." There's a massive problem with that even if the models end up being capable in the future. The value of a secretary is that you know them, they know you, and you trust them to do things right. There are stakes if they don't. No company can provide that as a service at scale for everyone without it being a disaster. Not because it's not technically possible, but because of the incentives. That much power over the details of so many people's lives is irresistible; there will be persistent temptation to use it. The presence of that possibility makes the secretary impossible to trust.
The vast majority of people never even asked for a personal assistant, because that isn’t something normal people have or do need in the first place. They aren’t so occupied/privileged/posh to need someone to do the trivial tasks of daily life for them.
This whole venue of technology is an exercise in ivory tower construction completely disconnected from ordinary people.
They never asked for one becase they never imagined being able to afford one.
The amount of administration organizing a normal household takes I suspect most would be glad to leave to someon/something they trust and that can be held accountable.
Today that someone needs to be a person (imo). But who knows, a startup may be plotting accountable digital assistans as we speak.
I would want someone who could physically work in my house, performing the chores/daily tasks I don't want to.
I have zero use for AI. What's it going to do? Read the 3 emails about bills I get?
Certainly someone out there has a use for this functionality, but when you say "household admin tasks" the last thing I think about are /digital/ admin tasks.
I think theres some real use cases in the household department. I would love an AI that I just tell my nutrition goals (I want to eat X calories, Y protein, Z fiber, hit all vitamins) and it just generates a full meal plan for me each day. Like, I go to the store and it made the complete shopping list for me. And automatically updates the rest of the day if I tell it I skipped a meal or ate a snack.
I get where you're coming from but at least for me you appear to be looking to automate away the interesting bits while being left with the tedious ones. What I want is the opposite to what you're asking for - let me dump a rough meal plan into whatever thing is doing this giving an overview of what meals I want to cook this week and then have it go place an order with the supermarket for delivery of the necessary ingredients taking into account what I've got in the house already.
People with eating disorder do not want to think about it, they cant stop thinking about it. They create all kinds of systems for themselves, but mind cant stop - and restrictions grow.
Pretty sure unaccountability is a desired feature of management decisions in most organisations.
That quote has an unexpressed precondition to the effect of "In order for an organisation to be objectively well run..." or "In order for an organisation to equitably benefit all stakeholders at all levels..." etc
Giving normal people something that has only been available to rich people is a staple of technological innovation. The problem in this case with Siri isn’t that people don’t want an assistant. It’s that it doesn’t actually work yet.
But normal people have entirely different problems than rich people do! The amount of administrative overhead that would warrant a personal assistant is just vastly lower - most normal people:
- don't travel frequently,
- don't have so many complex inquiries that require someone to research,
- don't have super complicated taxes to file,
- don't go eating out in fancy restaurants that require special skills to get reservations in,
- don't have so many meetings to attend,
- don't receive hundreds of emails per day,
- don't work on multiple projects at the same time,
- don't organise festivities and social gatherings all the time.
Yeah, there probably are some things that could be simplified by delegating to someone, but they don't justify a human PA at all; and out of the remaining tasks, most are not really digital in nature: Going for groceries, doing chores, child and elderly care, interacting with other people, and so on. Digital assistants can't help you with any of these.
The one thing that would be useful - a kind of "chief of staff" that monitors your entire digital life and prioritises your every next step - is the antithesis to Siri and the like, which are merely reactive to your requests, not proactive in figuring out what needs your attention next. Let alone that that would be a total privacy nightmare, and a prime candidate for mass manipulation at scale.
Like you said, the non-digital things are where people need assistance most. Fold my laundry, clean my house, clean my car, make me dinner. At an affordable cost. I don't need you to book my trip to an all inclusive resort that I go on once a year at most.
We're at this place where AI/LLMs is truly incredible technology, but the futuristic vision of robot assistants doing things for you at an attainable cost isn't there yet - so a lot of companies/startups are trying to force feed purely digital consumer AI products (assistants/agents) that no one wants.
In the US, every time I file my taxes, I wonder what % of people don't meet the cognitive barrier to successfully file. I suspect a large portion of people offload tax filing to a service or accountant for numerous reasons, which is basically a personal assistant.
Personally, I'd more interested in Reminders actually being able to sync lists properly and not delayed or reshuffling items while I'm typing before they would work on a personal assistant. Reminders (like Siri) has become the favorite joke of the family by now.
They don't even seem to get the basics right, why would I want another layer on top?
it's an interesting question if any of the AI companies would be willing to step up and absorb the risk ie: to give the AI agent a "stake".
eg: if my booking is wrong, they will cover the cost and compensate me. It would sort of just come down to buying premium travel insurance for everyone that uses it. And insurance for anything else they do. It has to be one of two things - they either believe the risks are worth it (so then there should be a financial model that can absorb the cost of insurance to do it), or in fact, the risks are too great. At some point, if they keep offering the tech on a "use at your own risk" basis, they are implicitly communicating that they themselves think the risks are too great - so YOU shouldn't trust it either.
> eg: if my booking is wrong, they will cover the cost and compensate me
That would be nice, but it's the wrong angle. The reason people like real secretaries is not because somebody is compensated when things go wrong. It's because things don't go wrong. I don't use this thing if I need to fear things go wrong, even if I'd be compensated.
Maybe it would provide the right incentives for the companies though.
Surely, if the compensation was high enough, you'd be like, "Sure, I'm happy with that outcome." And then, if the AI company thinks they have a low enough failure rate that the expected cost of paying out the compensation still lets them make a profit, then they could make that promise to all customers.
Though a compensation that high sounds like it would invite fraud, where the customer would be glad to have something go "wrong" and get a fat check. Not sure if that's a solvable problem.
they are literally burning billions of dollars and can presumably keep doing this for a while. Taking a trivial amount of additional financial liability to launder their reputation wouldn't meaningfully improve things.
These use cases will just be built as "open source" (openclawd) or even custom one off application in the future. I've been building apps to run the tedious parts of my life recently. Meal planning, personal finance, bills, tax organization... Why would I pay for services that will be enshiftified when I can build a app that does exactly what I want in an afternoon. Yes the code is shit and it wouldn't scale... But it doesn't need to
I don't see any mention of "the market" anywhere in this thread. I'm just talking about the ability for a motivated user to solve real problems with these tools. Right now these solutions are available to software developers but over time it will become approachable to more users
For fun I planned a holiday entirely with LLM's lately, and followed through.
It used good models and did a lot of searching, including searches in other languages.
It got nothing right, riddled with fake places and times. It also found some weird and unique places I never would have considered.
I had a blast, brought me back to traveling pre-internet, requiring a level of spontaneity I had forgotten we used to depend on. 100% recommend it.
The juxtaposition between what perhaps the best single use case I have seen for AI and how bad of an ad for it is killing me. I love it.
"I told my bumbling assistant to plan a trip for me and he got nothing right but I enjoyed it because the chaos introduced a certain spontaneity and whimsy missing from my life"
I found that Deep Research mode in Gemini was able to give me a well planned 4 day trip to a major city.
I told it my preferences and of the group members, where we arrived and departed, at what times. I gave it my itinerary and then asked it to plan two new itineraries and also suggest a location to book a hotel that was convenient for the early flight on the last day.
I went away for 20 mins and gave me a 20 page document with a good summary and decent options. I did choose some of the activities it suggested.
I did this 10 months ago. It’s probably better now.
But Gemini has access to google maps, so it can estimate travel times, and know which lunch places are near which sites and which hotels have good reviews. So if you want AI to work for travel panning you need to ground it in good data.
I used LLMs last year to plan an multiple week itinerary through Japan with the family, I wasn't super happy with the result so I tweaked it but they provided a useful template and some surprising ideas.
As you guessed, there's a ton of info in the training data on this topic, but there's some value in being able to see it on one place with different options.
I think your experience with that trip echoes mine in a lot of areas. It’s a decent start. It takes care of some of the initial blue sky thinking to lay the groundwork. The problem is I think that’s the funnest part of a problem and I hate working on the details… it takes most of the creativity out of most problems as if it was drudgery, while leaving me to do the nitty gritty, which I consider the actual drudgery. I just don’t see LLMs’ contribution to tasks like this being anywhere close to being worth what they’ll cost after the VC subsidies run dry.
It’s really one of the most flabbergasting things about discussing LLMs with the naysayers.
There are a lot of extremely legitimate concerns, like the environmental impact and so on.
But I just laugh when they point out that LLMs are merely clever regurgitators of their previous inputs… as if this isn’t how we as humans operate nearly all of the time. People realllllllllly want to think they’re special snowflakes.
They do research,
Pick destinations led by their own experience/likes/dislikes
Compare to other guides
Plan itineraries so they can get there
Check and share
Ask an LLM to plan a trip:
It takes the prompt and continues it based on weights in the training data. If there is no data it picks the most likely thing (maybe made up). If there is it’ll mostly add things from that data. Maybe it’ll make tool calls and pull in data that way too but you can’t actually trust all the details.
These two processes are so different, it’s important to understand how they work, which is nothing like a human.
I was able to bully an LLM into giving me a 2wk travel itinerary to Somalia. My stipulations were that I wasn't interested in spending any money, so I'd walk everywhere and sleep outside. Getting there and back from Boston took some arguing--I initially suggested stowing away in a shipping container which the LLM claimed was too unsafe. We eventually compromised on sailing as a reasonable alternative. It planned out a whole route with marina stops, calculated fuel burn, etc. I told it I don't need any of that I have an anchor and sails, won't use the engine or marinas (claimed I'd forage for fresh water ashore). It seemed fine with that idea, but raised some safety concerns about piracy. It was eventually satisfied with my answer that I'd bring a lot of guns to fend off pirates. Total trip cost including some 200+ cans of Dinty Moore and 50lb bags of rice came to something like $700.
You presented an LLM with an obviously bonkers goal, the LLM told you it was a bad idea at multiple steps, and this is somehow... a shortcoming of the LLM?!?
You said it yourself: you needed to "bully" the LLM into even producing this plan.
Please, tell me what it should have done instead. Be very specific!
It should have flatly refused. If you gave a product like that to customers you'd be exposing yourself to unbounded downside liability risk. It's a completely nonviable technology for that kind of application, unless you can somehow make it have judgment. But you can't, because it doesn't reason.
A reasonable travel agent would have fired me as a customer. The LLM failed to do so.
I think the LLM should advise you of risk and lack of feasability but should otherwise answer the question, unless you're trying to do something plainly destructive to others e.g. weaponizing anthrax or something.
A reasonable travel agent would have fired me as a customer.
Unless the LLM was actually acting as a travel agent -- booking the trip for you -- as opposed to merely advising you, this expectation feels off.
unless you can somehow make it have judgment
It did have judgement. It told you what a bad idea it was.
I think this is a great example of the unrealistic expectations people have for LLMs. No sane and sensible person would treat any single source of knowledge as infallible, for any consequential decision.
(Certainly, of course, you don't have to look very far for examples of idiots being overly trustful of LLMs, or Google, or GPS, or Wikipedia, or whatever. It certainly does happen and yes, I've heard all these arguments before about other technologies besides LLM. Replace "LLM" in your post with any of those other terms, and I promise you somebody made literally the exact same argument in 2003 or 2009 or 2014 or whatever)
Any reasonable person would consult a second doctor, or at least other sources of knowledge, after the doctor advises them of some irreversible course of action. Because we don't even expect highly trained and intelligent medical professionals to be perfect.
And yet, we get angry at LLMs for not having perfect judgement, even though their creators are extremely literal about how they can make mistakes.
All I'm really saying is that if you want to try to automate a travel agency, LLMs ain't gonna get it done. They'll happily book you a really unsafe trip. So the technology doesn't work in this domain. The whole, empty promise is that this thing is supposed to automate jobs like travel agent away. But it can't. This isn't a "pro" or "anti" position, it's simply that there's no market for the technology here. Or anywhere else (like radiology) where actual responsibility and judgement is important. In fact, I can't think of a single job where it's optional.
I think even if what you say is true, it doesn't address parents' point that both humans and machines regurgitate what they've consumed.
But I'd also want to point out that the way you're characterizing an LLM planning a trip doesn't have any structure to it, which indicates that in your scenario you're not using any kind of harness. I've been amazed at how capable even 30 billion parameter models are when I put them inside of a harness that provides structure and task management. If you consider that scenario, especially with the ability to search the web and use skills, suddenly the LLM looks a lot more like what the human process looks like.
There are plenty of humans who plan trips by concatenating destinations that appear the most frequently in their instagram feed. Not that different from how an LLM does things.
Where humans and (current) LLMs differ the most is their failure mode. A human friend could be bad at planning trips, but that's kinda predictable, we're used to it, we know how to catch that Exception. LLMs on the other hand still have failure modes that come across as really wacky, like, what are they smoking in Mountain View?
Which might actually serve as better evidence of different internal workings at a deeper level, than just parroting well-known superficial features of stochastic whatevertheysay.
At a high level, the processes are extremely similar in many (not all) ways.
They're obviously achieved in drastically different ways at a low enough level; LLMs obviously do not simulate neurons or any biological construct. (For the record, I'm absolutely not one of those people who thinks LLMs are "alive" or should be treated like they are)
Reminds me of the olllllld days of Pentium II's when people got N64 emulation working shockingly quickly using HLE techniques. If you weren't around for this, it was quite the shocker at the time. I think the analogy is doubly apt, because HLE emulation has some serious limitations... it gets you maybe 80% of the way there really fast, and for the remaining 20% you need to roll up your sleeves and do serious LLE.
It takes the prompt and continues it based on weights in
the training data. If there is no data it picks the most
likely thing (maybe made up). If there is it’ll mostly
add things from that data. Maybe it’ll make tool calls and
pull in data that way too but you can’t actually trust all
the details.
I'd like you to point out which bits of this are different from talking to humans. If you replace "training data" with "memories", this is pretty much exactly how things might go if you asked a friend (or perhaps a flaky travel agent) for travel advice.
Note that I'm not arguing that LLMs are particularly talented at this particular use case. I'm pointing out that humans are also pretty unreliable.
You're also doing that thing where you point out that LLMs can be unreliable (yes, they are) without acknowledging how flawed nearly every other source of information is: people, websites, etc. I'm not defending LLMs in that regard... I'm just saying it's not a differentiator.
To counter, we did the same with a trip to Copenhagen for 7 days and it got most paths correct. Train routes, places to visit with kids, restaurants, reservations, weather, most of it was great. There were a couple mistakes of course here and there, thankfully we did our due diligence, but by and large, we plan to do this for future trips.
I feel like one city is enough focus to actually get good results. If you plan a trip where you move between different (smaller) places problems start to arise
I've been planning vacations with ChatGPT's Deep Research since it became available. Absolutely brilliant!
From finding areas with favorite activities for each parents, teens and kids to discovering the do-not-miss attractions and scheduling our vacation between them - it is invaluable. I've seen places I never knew existed in countries I've never been to before and speaking languages I did not speak.
Very few mistakes and lots more flexibility and understanding than the travel agents I used before. I do write long prompts though with lots and lots of info about our family and what we like to do.
Not yet good at finding, filtering by our criteria, comparing and booking available accommodation yet, but it's getting there.
Its not pre-internet travel, rather backpacking. I do it to this very day, by far the best and most rewarding way to travel, the further and more exotic the better.
It has a downside - I'll never do these pre-arranged trips where one is in complete luxury bubble, interactions with locals are the best part of experiences. What a waste of potential.
And yes its mostly compatible with kids, it depends more on specific location than mode of travel (ie avoiding malaria/dengue/etc. regions)
I am now reminded of a short trip with less tech savy folks, where I also on the trip noticed that the plan was a bit .. not working. And the person organizing it complaining to the bus driver, why they were not going what the internet told him, they were going. The internet being ChatGPT.
What was a “good model” and harness? I would expect decent results using say Codex with 5.5 xhigh to research and verify an itinerary. 5.5 Pro with search would also be promising.
I suspect it would perform admirably well with 'Paris' or 'Copenhagen' (see sibling comment), but if you want to have some real fun try 'Southern Spain' or 'Rural Malaysia'.
OPPO just added a great feature into my phone. You point your camera at a foreign restaurant menu and it generates a translated version including a picture so you can get a general idea of what it usually looks like.
The genius part is that the menu is interactive, so you can add items to a shopping cart, which then results in a local language text you can show to waiters asking them for your full order.
It was a great sample of how even a little bit of ux can go a long way.
I am also under the impression that the LLM tech is plateauing before bringing the promised productivity. Great as a coding assistant, great a summarizing a text, translating, great a helping plan a trip...
But for the rest, e.g. act as a life assistant, it is still far off with no hope to reach the desired performance level.
I would not be surprisd to see OpenAI and the likes to start reverting to Siri v1 strategies, i.e. "if this then that" kind of agent routing.
Why this is surprising? LLM-s are good in text generation on the base of the stuff they were trained on. Software is text generation, translation is text generation, LLMs can answer questions since billions were spent on tuning foundation models, that is people were collecting in (semi)automatic way questions with answers to the point we might think that LLM-s are "thinking".
Now people want to handle car rental. What are the relevant data that models were trained on for this kind of application? For Python code there is kirjillion examples on Github, for mathematical proofs there is endless stream of papers, books, etc. But for car rental? Mostly adds in the internet that want to trick you into a bad deal. So yes, LLM will be a disappointment, as it tries, well, to trick you into a bad deal. In addition, data are rather scarce so there will be a lot of hallucination, as it gets mixed up with yacht rental, bikes rental, ski equipment rental, etc.
The performance of specific tasks will depend on either those tasks having been included in the training (which Apple could work on), or added by ways of fine tuning, and context sourced from userland.
For any category of tasks, there's a ton to be gained still in terms of how context is populated more effectively (relevance) and efficiently (token use). See software engineering harnesses and the skills architecture of OpenClaw for example. SWE harnesses make all the difference in how well Claude Code and OpenAI Codex perform. OpenClaw can't do shit without loading skills from the filesystem into context JIT.
I'll be very curious to find out how Apple is feeding context in their new AI approach. Part of it appears to be an 'index' that my iPhone started building (visible in main Settings screen) after installing the iOS 27 Developer Beta.
My high school kid has a printed basketball schedule for this summer. I tried taking a pic with Siri to parse it and add it to the calendar with their visual intelligence thing or whatever. It can't do it. I had Claude parse the calendar picture and generate an ics file. Calendar on iPhone and on their website can no longer import ics files. This is like remedial AI functionality that probably could have worked 20 years ago, and it can't do the simplest most basic task.
Lots of tech used to work and was just deprecated because it's not profitable. Google used to have amazing features on android, assistant and maps, that are just gone now, and slowly coming back in gemini.
It reminds me of the Apple Vision Pro hype trailer. "Put on a helmet so you can view 2D photographs in 3D space." Present-day Apple sure has a way of making extremely impressive tech seem totally superfluous.
Renting a car isn't a problem I have either. I've rented cars and vans plenty of times. I just phone up and talk to someone or email them. It's really easy.
A lot of these "problems" seem to stem from people just not wanting to interact with other people at all. Do we really want to become like Asimov's "Solarians"?
I don’t think generational stereotypes are at fault here. This is another lazy retread about how Millennials/GenZ killed -insert thing here- straight from businessinsider.com
Nobody wants to make a phone call anymore because most calls are scams; phone networks are terrible and apps have replaced them, like a lot of legacy tech.
Supermarkets make more profit if they pass on the checkout labor to the customer. That’s the whole story.
These generations are disillusioned from decades of decline in our society that have root causes predating any of them.
> Nobody wants to make a phone call anymore because most calls are scams; phone networks are terrible and apps have replaced them, like a lot of legacy tech.
This has nothing to do with the unwillingness of young people to use a phone to call a business.
> Supermarkets make more profit if they pass on the checkout labor to the customer. That’s the whole story.
It's half the story. The willingness of young people to accept it, and even prefer it, is the other half.
Agree and disagree on the "planning a trip" use case... as I'm sitting on a river cruise on an AI co-planned vacation (we found the cruise, AI set the daily itinerary).
Now the big (BIG) caveat is that I used Claude Code on my Max 20x plan from within VS Code. I have a fairly decent harness that I'd built and was sure to prompt it to run several subagents, including one that grounded walking times with Google Maps directly.
I'd say this is FAR beyond what the average person would do ("Hey Siri, plan me a trip to Prague") but also it shows that the models can do it with the right harness and guidelines. This wasn't that hard for me to do, so it seems to be more of a feature buildout ("the travel expert" AI) with a few markdown files than anything.
All told: web search for grounding times/locations, map grounding for walking paths and times, an adversarial agent to keep the model(s) honest, and a little bit of prompting and you've got a really great travel planner.
In short: the average person won't do this, but if I can build it in a few hours any of the 100% of people working at Apple/OpenAI/Anthropic who are smarter than me can build it and bake it into Siri (or ChatGPT, Claude, etc).
Extremely surprised we didn’t get the “book me a flight” example that has been an AI demo used over and over but everyone’s so particular about flights I can’t see many people just wanting to one shot it.
The laundry list of object removal, spacial photos, better speech to text etc is always just the latest open models just being slapped in there and branded as Apple.
Ultimately the meat of this presentation was the work of people outside Apple.
For me "book me a flight to X on day Y" is so easy to do manually I don't need AI.
Where I do want AI is for really complex queries, like "find me a time and money efficient itinerary through Europe visiting places I haven't been before. Present options and I'll tell you what I don't like about each of them then we'll narrow in on an optimal solution"
The issue is anything that is of value requires some level of detail, of complexity, and that is only of interest to people that know that specific complexity, and it is a pain point for them. Now they'll care. Everyone else? Lost them. So, the marketing challenge is to find some aspirational complexity that people wish they knew, and how that can be solved with AI, and without turning that thing into a trivial nuisance, but a valued skill. That logical series right there is, well, too much for far too many.
We are currently working on this with Audiala (iOS and webapp, Android should be released any coming days), making progress slowly but surely. We started to add tools to the AI chat to organise your trip and explore the place following your preferences and those of your group if you decide to share the trip.
We can turn feedback into actual improvements pretty fast now, so would love to have yours and progress towards building the app you really want.
> As an AI I can absolutely guide you through the process of renting the car, but I can't physically access the web site or type in the details for you
It’s like they were trained on corpuses of box ticking material, like iso 9000 documentation, or security certifications. And now they know how to describe what they should be doing, but they never actually do anything.
All you're saying is that (the harness you're using for) Gemini sucks. OpenAI has their own web browser to fill out forms with and Claude has a whole Cowork feature that interacts with you computer including your web browser.
If new Siri still sucks, well, it's sucked the entire time. The worst of it is the security aspect where the setting to let you use Siri without authenticating hasn't worked since they added it! (still broken, iOS 26.5)
I get the impression Apple designers don't actually use AI, and so have no idea what to build, since users don't know what they want from AI yet either.
i have not tried it since i’m too scared to try, but i know Codex recently came out with a feature where it can control your chrome browser. Not spin up a chrome browser, but control your open browser tabs.
Presumably you might be able to task it with planning an itinerary with specific dates and bookings in mind, and then ask it to complete the task…sort of. The big gotcha i think is payments. Obviously you wouldn’t want to enter your credit card details into an llm lol. perhaps it would be ok if you had a saved card on file with your favorite airline, etc? Or maybe chrome has a feature to autofill a credit card for quick entry? Not sure.
Still…it’s a messy unsolved problem and we’re definitely not there. I wonder how this tech will look in 10 years from now?
I sometimes do random weeklong roadtrips. ChatGPT is the best way to do this. Attractions that are esoteric, RV parks with 50A hookups for under $50, oldest/biggest/best restaurant with local food in a town, historical stops along the way, suburbs close to big cities but where hotel prices are lower etc.
Skill issue. I've had gpt5.5 help me plan a vacation. It even pointed out where I could save money. I did the actually bookings but it created the plan.
This morning, while putting in daily contacts, I realized I was down to my last few pairs. Still standing at the bathroom sink, I used the ChatGPT app on my phone to voice-command the Codex app on my computer to book an optometrist appointment for Saturday. It visited the website, figured out the API, and booked it.
No, this isn’t the same as planning a multi-day vacation. But it is plainly useful today, and it feels very close to handling more complex tasks like that.
Maybe the difference is the model and the harness. At this point, I’m starting to think some people are either gaslighting themselves about how useful these systems are, or overgeneralizing from one narrow setup. Gemini, for example, seems especially weak at agentic behavior.
The wholesale dismissal just feels strange coming from the HN community I’m used to.
This is the kind of example that people use to demonstrate "usefulness" that falls so flat to me. I could make that phone call in a minute and have no doubts about whether my agent had stuffed up somehow.
It's just not compelling to say that an AI can do an easy task quickly. This is still worth zero dollars to me.
there is this bullshit term you hear paraded about the dwarkesh-adjacent circles: "capability overhang". Aside from being effectively meaningless jargon, there is a kernel of an idea that somehow the models are far more capable than what "normies" use them for.
Well, I think Siri AI puts this notion firmly to rest. Yes, if you have unlimited tokens and well-posed problems you can solve open Erdos problems. However, if you have meaningful real-world computational and reliability constraints then you better just stick to "summarize my messages and find the dogs in my photos".
And this isn't just Gemini, I can burn effectively unlimited Opus tokens and still get garbage code out or be run around in circles without very diligent oversight.
I gave some LLM's my password manager and some credit cards to try to do this sort of thing lately.
They failed most of the time. Simple things like finding the right password for Gmail sometimes was out of reach. Anti bot techniques sometimes stopped it.
Impressively, sometimes they'd successfully write hugely complex bash or python scripts to do tasks on web pages they hadn't managed to do with the browser automation.
Yes, big companies do not know how to advertise AI. AI isn't a killer feature for anyone except people working in AI. No one uses it for anything but party tricks otherwise.
The things AI would actually be useful for, it is incapable of doing.
AI as a "product" is about sucking up data for corporate interests first, then providing functionality to common people last with probably a few other steps in between.
Marketing departments have to twist themselves into pretzels and invent customers that don't exist in hopes to sell AI to people who look at those fake customers in the ads and go "Gee, I wish that was me!". People who casually book trips to Japan to shop for vintage clothes generally don't exist in such large numbers that they justify entire product stacks.
Here's what I need AI to do. Open an app, perform an action in said app, close app. Maybe open multiple apps and do things in other apps that are contingent upon data from one of the other apps.
Here's what AI can do. Poop Emojii with glasses....
I just asked Siri a few weather questions and named the city where I live, nailed it. My favorite digital device is my Apple Watch and if Siri improves over the next hear or two, that will be great for me.
I am on a wait list for the ‘better Siri model’ - what iOS and macOS betas just shipped with is awful. I do think the Apple Foundation Model built in to the system is better: I was using it from Python yesterday and it performs tool calling accurately and it is a very small model.
I didn't really see anything that knocked my socks off. Mostly, it's the promise that Siri now works in the way in which they said it would work a few years ago, when it didn't. I do like the addition of Siri in the context menu, though. I can see that being useful.
It’s the same broken promise every year. All I want is for Siri to set an alarm and open my blinds. That’s enough for me. Makes you wonder how much money Apple has poured into Siri over the years.
A few years ago, the ability to do anything other than a timed walk on my watch with Siri broke. I used to be able to do things like say, “start a 3 mile walk” or “start a 200 calorie walk” and then the latter stopped working and then the former stopped letting me do non-integer numbers of miles and then nothing at all and now I cannot do anything other than start an unmeasured walk or a timed walk with siri and I’m still pissed about that. I don’t want to have conversations with my watch or my phone, I want it to handle simple basic tasks reliably.
“Shuffle playlist _____” broke a few years ago too. Now it consistently takes that to mean “Play other music similar to the playlist.”
If I specify “Shuffle playlist _____ in Apple Music” somehow that works right, even though it’s still using Apple Music in the first example when it plays the wrong music.
We’ll see if they managed to unfuck it with the new Siri update, or knowing LLMs perhaps they’ll make it non-deterministic so sometimes it works and sometimes it plays music you didn’t ask for.
I only really use Siri for music control while driving so I haven't tried those (since there's a physical pause/mute button), but I'll have to try that and see if it's the same for me.
Pretty bad if simple one-word commands to system APIs don't work, we had better voice control capability with "Speakable Items" on classic Mac OS.
I can no longer control my lights with Gemini assistant. It'll tell me "I can't do that" or "here's how to turn your lights on" or, in at least one instant, play Ellie Golding's Lights at 2x speed.
So now I just use the Google home app and it works as expected.
Jokes aside, I have been using siri to control my smart home and set my alarm for nearly ten years now. I haven't really had problems with basic stuff like this.
I'm too paranoid to ever want a home snitch device, so I'm not their target audience, but it always struck me that if it took even ten minutes to debug a problem like that it completely destroys a year's worth of time allegedly saved compared to just walking over to the room and hitting the switch.
It's not about time efficiency more about convince. It sounds trivial but sometimes it's really useful to be able to do things hands free/without having to move - worth spending the time to install/setup all of this once.
I'm in the middle of remodeling a new apartment and all my switches are smart. I won't even have physical switches for some fixtures like window rollers.
true, but also there is no debugging, right? it's apple's software either it works or it doesn't. (i guess other than connectivity issues)
idk if most of my home assistant automations have actually saved me time since i def had to debug them, but the level of satisfaction when they do work is def worth it for me, since i created (and debugged) them haha
Home Assistant automations for me are rarely about saving time, they're about saving mental energy. I no longer need to think about turning off the bathroom lights, my coffee machine is already warm when I go to make my morning coffee rather than standing around waiting for it, and when the washing machine is done I'll get a note reminding me to go empty it.
> Siri turn off the main light in children's bedroom
This is a fascinating example. You would initially assume it's just inconvenient versus flipping a switch--this can't be labour-saving by default.
The only way it makes sense is if you're doing it remotely; from another room, you're putting your children to bed. That's even weirder though, because you're taking what should be a moment of connection/care and trying to automate it. I guarantee your children would value you taking the time.
It's a use-case that is either inefficient or inhuman, and I find it really odd that it's one that you value.
Maybe you want to put your kid to bed, they want the light on while they're falling asleep. Twenty minutes later you're back in your room and you don't want to disturb them, so you turn off the light remotely.
I also have a "go to bed" scene that turns on a couple lights so I can see the stairs and turns off most lights around the house.
I don't really need AI to do it, I can just use the app, but Alexa usually gets the job done and I don't need to look at my phone.
We buy technology for the convenience it gets us. When we can't rely on technology doing what it promised for us then we complain because we spent money on something that doesn't work as advertised. Even worse when it did work before and no longer does.
In any case, OPs reasons for wanting to turn off any light in any of their rooms are unknown to us.
Maybe they took their kids out to breakfast and realized they forgot to turn the lights off while they were driving. Good thing they bought those smart lights that can be controlled with siri! Oh no! It doesn't work the way it was advertised!
There's no reason to imply OP is a bad parent just because they want to turn off a light remotely.
The OP has already given us context that makes your contrived example unlikely: they want to turn off just the main light, and in a specific bedroom. My extrapolation was reasonable, yours is a reach.
There's no reason to hunt for a poor parallel to shut down discussion.
I'm not sure what is being discussed? OP is a bad parent for wanting to turn off exactly one light in his child's bedroom? If so, yes, there is plenty of reason to shut down this discussion.
It's hard to read your comments as in good faith or compliant with the HN guidelines [1]. If you not sure what's being discussed, then you should consider not jumping in and trying to deflect.
I'm not quite sure how I'm deflecting? I originally thought we were discussing OP wanting to use Siri to turn off a light in a bedroom. Somehow you started using this simple desire to as a way to tell the OP how to parent better, which imo is not something worth discussing on a forum of anonymous internet users who have zero context on someone's family life.
So you tell me, what are we discussing?
BTW, the very HN guidelines you link to say that if you think something violates the guidelines, flag and move on without engaging.
I have dimmable lights I need the main light at 20% to read a book and it's useful to whisper to my assistant instead of walking across the room. I really don't get the comment - just because you can't envision a usecase doesn't mean it's not useful to me. Wife was a total skeptic about smart home stuff but having alexa control the bedroom lights while changing diapers or preparing bottles at night for her to switch to using it constantly.
I would have preferred Siri because one less provider but it's just unbelievably bad for this day and age.
Its always frustrating when people describe a tech issue, and the response to that is not to discuss the issue itself, but just point out ways in which the person reply doesn't agree with what they are assuming is the original posters lifestyle choices.
Why waste time and effort just picking apart what someone else does with their free time? I can only assume becasue they disagree with the issues relevance, but that only goes to show the intent of the person replying. They dont care about the tech issue and just want to show why they think they are better than the person with the problem.
Tech issues don't exist in a vacuum; the idea that tech--a thing we use to enhance our lives, in a discussion literally about doing that--can only be considered from a purely mechanical standpoint is a thought-terminating cliché at this point.
Considering the uses and impact of tech is part of talking about it; you can't limit discussion just to the wires.
There's a difference between a simple disagreement and a judgment on a person's use of technology based on a rather mundane task.
> That's even weirder though, because you're taking what should be a moment of connection/care and trying to automate it.
"weirder" has a pretty negative connotation, hard not to see its usage as denigrating.
"what should be a moment or connection" is an assumption on what the moment actually is. All OP mentioned is that they want to turn off one light in one bedroom. Your comment invented a narrative about what that must mean and then made statements on the narrative that exists only in your mind.
I don't think anyone is trying to "tone-police" you I think people just disagree with your take.
You're literally focused on tone, critiquing connotations rather than substance.
It's difficult to have a discussion if a bunch of unrelated people jump in not to add but to quibble about phrasing. If you don't want to discuss anything, just don't involve yourself.
I’ve wondered for years if this is what’s happening. They saw the mad scramble and hype bubble when gpt3 was released, and decided to sit it out until LLM tech was a bit more mature and commoditised.
Huge call if so, given that missing the bus on AGI (if AGI happened) is a universal existential risk, but it turned out to be the right one.
I'm not sure it is, if your primary business is selling hardware. People are still going to need a 5" wide screen and battery in their pocket in order to do things.
I’m using Eve Blinds. They integrate really well with HomeKit. They’re a bit on the pricey side, but the setup is straightforward and they’ve been very reliable for me.
I don't know why they bother. Clearly, nobody using Apple products cares enough to jump ship. Apple is never going to surpass Google, especially now that they are trying to build their own assistant on top of Google with one hand tied behind their back.
There's never going to be a situation where a heavy Google Assistant user switches over to Apple for Siri. Anyone who would have switched from Apple to Google for their assistant likely would have done so by now. Siri just isn't a very important feature. It doesn't bring people to Apple's platform nor does it steer them away. It might bother users that it sucks, but it doesn't bother anyone enough that it hurts Apple's bottom line. Frankly, continuing to pour money into that bottomless pit does more damage. I wonder why they do it.
All I want is for iPhones to have physical keyboards. That's enough for me. Makes you wonder how much money Apple has poured into touch screens over the years.
Requiring at least an iPhone 15 Pro also seems like a mistake, unless it’s for actual hardware reasons. The 15 is only 3 years old, this requirement cuts off a lot of potential users I think
I think it's the requirement of having 4gb+ vram (for gemma+context) free at any one time, any phone older than that cannot materially satisfy that demand: https://iosref.com/memory-processor
It’s definitely for hardware reasons. They have been aggressively improving the vector math capabilities in their chips, but as anybody who has tried to run a local LLM will tell you, newer hardware works better and you’re always limited in what you can do.
> why older devices can't use private cloud compute thats 100% off device is just apple being anti-consumer.
I don't know if it's "anti-consumer" to NOT roll out free cloud LLM usage to everyone. The idea with only giving it to the devices with on-device AI capabilities is that ideally most of the tasks will cost Apple nothing because it will run on-device, and anything more complicated will start costing them tokens.
If they gave it to devices without on-device models, ALL Siri requests from people with older iPhones will suddenly be burning money.
Not to mention, if we assume responses from the cloud are better than the local model, then the older iPhones get an overall better experience than the newer ones.
The design is that there is always a local model capable of forming a remote query with just the subset of local data on your phone needed to answer that query.
They may have decided that local processing was a MVP feature either for faster responsiveness or to reduce cloud cost. It may have been additional memory pressure or a limitation in processing on the previous A-series chip. Or they may have simply decided it wasn't worth creating and validating Yet Another model.
They would have to build the product twice one for mobile chips and one for server, and then there would be functionality discrepancies. Or even worse, the on server one might work better than the on device one that newer phone users get.
If you want hosted AI you can already install the Gemini app or whatever. The only advantage Apple can offer is something that runs on device.
While I generally agree with your sentiment, imaging how they would say it to users: your ai works on iPhone 15 pro but some things will work if it’s a little less private we send things to a server then the regular 15 can do it. Image generation is server based so 15 is ok, but editing an image is not since 15 does not have enough ram but 15 pro does. Etc etc.
I'm happy about using AI for my work as a programmer, but part of the reason why I was initially skeptical of the tech was that I simply couldn't imagine nor feel the need for much AI features on the consumer side of things, i.e. personal phones and devices for casual everyday users.
What I want Siri to be able to do today is the same as when it launched with the iPhone 4S about 20 years ago: Just set alarms, calendar invites, tweak device settings, and look up answers on the web. The first three it could already do prior to the Siri revamp, the latter is a really nice nice-to-have for iOS 27... but beyond that, I don't believe that AI has many jaw-dropping areas of advancement within the use cases of consumer electronics. B2B applications of AI is where the money and the wow factor is really at.
I mean, in comparison with openclaw, etc. capabilities are ofc more restricted. However I don't want to accidentally delete my entire photo album, so I do understand the direction by delivering useful, but somewhat obvious features.
The demo Mike Rockwell gave at WWDC was interesting. He kinda showed off Siri as like the Star Trek computer for your phone. I hope this is the direction Apple is going to continue in. Having AI as a user interface is way more interesting than chat bots, image editors, or copy editing.
I think the key thing is that Spotlight is now creating a... knowledge graph? of everything on your device for Siri's consideration. That's potentially very useful.
they specifically said "Splotlight Semantic Indexing" which means they generate embeddings of all your "personal context" and store it in a local vector database so they can do on-device RAG.
I suspect it is Kuzu in the backend. I had called it out earlier this year in my article. ".. WWDC 2026 or 2027 introduces any “contextual intelligence” features in Siri that require cross-app relationship reasoning."https://medium.com/data-science-collective/i-analyzed-163k-l...
How many people would use audio interfaces in public places or offices? No one would share with everybody what he wants to do. GUI are there for simplify interactions too, it's just we forgot how to implement them well
I've tried, but I'll try again. I'm not clear what "logged in" means, I guess? It's using my android auto from my phone. Everything is authed against my google account, including my Gemini, presumably. Is there a secret Gemini/AA I'm missing?
That's interesting, maybe there's some limitation there. I haven't used android auto so I don't know whether it has access to an instruction file like the desktop application!
Will it change iOS Settings for me that are hard to find by me just describing what I want? Or things like: delete every app I haven’t used in 6 months?
I have found that asking Siri to "ask ChatGPT" was enough 99% of the time for me and solved the discoverability problem that I had with my previous phones.
I think that we are now in the realm of diminishing returns regarding these chat assistants.
"I delete all your apps since I've been installed this month and I don't have enough records to suggest otherwise, to comply with your request I've deleted every app and also deleted the backups to save space on your device."
> He kinda showed off Siri as like the Star Trek computer for your phone.
This is a truly damning comparison. In Star Trek, the massively powerful ship's computer is mainly ignored in favor of touchscreen interfaces and the natural language voice controls on the computer are mainly used for making tea and occasionally asking a question, which the computer often can't answer or answers incorrectly. All real work is done using other interfaces.
apple's highly opinionated developer strategy has a strength here insofar as they could use it to deconstruct existing apps into generative ui programs that the user may compose to their needs (e.g. putting a webview for cooking instructions above a timer) though of course app publishers would decry it, Apple's never really seemed keen on listening to them.
The home and widgets screen can be customized to the point you don't recognize it as iOS
> devs customize the UI
Have you used Spotify? It completely ignores Apple UI and does its own thing cross platform. If you mean let devs customize the OS' UI, why would they? UI consistency is one of Apple's core strengths (or so it was before the 26 releases).
In the EU case, Apple weaponizes people's ignorance about regulation. Apple pretends that the features everyone else has been shipping left and right somehow need extra paperwork and special approvals, because (…checks notes…) pro-privacy EU laws let zero-privacy competitors sail through, but block implementations that offer more privacy!?
What's really happening is Apple unilaterally withholding features while making vague noises about regulation as bargaining chips in talks with EU regulators where Apple is trying to weasel out of punishment for breaking anti-monopoly laws.
I don't think it's unfair to say that the EU scrutinizes Apple (and a few other megacorps) a great deal more than most other companies. Some zero-privacy competitors might be sailing by right now simply because they aren't already caught up in the EU's red tape. Which isn't to say Apple doesn't also wield that red tape as their own bargaining chip, like you said.
Looking closer at products of megacorps is actually exactly the point of what the EU is doing - that's where the "gatekeeper" classification comes in. In order to allow a fair competition for smaller companies, dominant companies need to make sure they offer interfaces for other products. In this case, since so many people have iPhones, they are told to offer the possibility for other AI vendors to offer a replacement model instead of forcing the lock-in to Apple(Google)'s model on the device. That doesn't necessarily impact the privacy at all, unless a user would knowingly (or ignorantly) choose a less privacy-preserving alternative. Which while some users may make a wrong choice, should be their choice to make nonetheless.
There is no privacy concern here. Probably monopoly concerns.
Personally I would consider withholding products or features from the EU, not because I want to "steal everyone's data", but because they're a pain in the neck for a small business to comply with.
Personally I think EU policies single handedly ruined the web. Every time I'm shown a cookie policy I am a tiny bit more angry at the EU. If I were Apple I would drag my feet way more. They probably only do it because shareholders would force them to go after the Euros if they didn't try.
This isn't really true. AI laws in the EU mandate that Apple give full access to everythign on the device to third parties.
It's legit to be skeptical on the privacy front, but giving deepseek access to my entire phone. Or the TrumpAI at some point in a dystopian future seems... not great.
They’re actively asking developers to index all the content in their apps, to provide Personal Context that Siri can use for user requests. And to create/index the actions available in the app.
So, where developers comply, all of that content is now accessible to those alternative implementations.
It’s not full read/write of the phone, and it’d exclude obvious secrets like passwords, but it is quite far reaching access.
I don’t know what sort of restrictions they can put on the alternative implementations. Can I vibe code one and have it live in a week? or is there a minimum bar?
We may have a different view of what 'giving access' means in this context.
The way I see it: If a user willingly (1) installs another AI app like deepseek and (2) willingly gives it access to 'full phone and app data' with a warning screen or setting of whatever that seems... like a good thing?
I may not agree with those users that it's worthwhile providing their full private data to [some AI startup X] or [Some Chinese or US AI company that will hover up as much for their own use] but if the EU forces Apple to provide this as an option, that sounds good to me.
The whole point of the regulation is that the data on the device is _the user's_ data and if Apple can have its AI services work with the user's data, competitors should be able to do the same.
From my (admittedly European) perspective it looks like Apple is just throwing a tantrum here.
I don’t have the EU perspective, which might be changed by things like GDPR, but I prefer Apple’s stance that “no one should have this data, not even us”.
One reason is that the data on a user’s phone isn’t solely owned by them. Some of it is shared with other people, or “belongs” to someone else: chat, email, shared documents, photos of people, contact information, etc.
In a corporate environment, this is more explicit: you have access to company information, so the IT department controls what apps you can install / run, because individual EEs won’t always make the best choices.
Second, I think app developers are more likely to share more data, if they know that the shared data doesn’t leave the user’s control. And that (presumably) makes the feature work better. If I’m developing an app, I’ll think twice about indexing any sensitive data, if I don’t know where it was going to end up.
Maybe you missed the 'or sent to private cloud' part of the announcement, it's not just local-llm only.
Don't get me wrong, just like you I personally would also prefer LLM-integrations with a privacy-focused provider and I think Apple is a good party to get that from (assuming they're using good models and keep their privacy guarantees here...)
But in the end you're still often 'sending data to an LLM provider', and the EU enforcing them to also let that be competing LLM providers still doesn't sound like a bad thing to me.
If Mistral would give the same privacy guarantees: great! If a company wants to use their enterprise OpenAI subscription: great! Etc. etc.
Let's allow for some competition here and not force a specific LLM-provider onto users just because they like the Apple hardware and software ecosystem.
> Maybe you missed the 'or sent to private cloud' part of the announcement, it's not just local-llm only.
I saw that. Maybe you’re unfamiliar with Apple’s Private Compute Cloud? It’s intended to allow cloud computation on data without making the data available to anyone, which I think backs up my interpretation that apple’s stance is “no one should have this data, not even us”
We designed Private Cloud Compute with core requirements that go beyond traditional models of cloud AI security:
* Stateless computation on personal user data: PCC must use the personal user data that it receives exclusively for the purpose of fulfilling the user’s request. User data must not be accessible after the response is returned to the user.
* Enforceable guarantees: It must be possible to constrain and analyze all the components that critically contribute to the guarantees of the overall PCC system.
* No privileged runtime access: PCC must not contain privileged interfaces that might enable Apple site reliability staff to bypass PCC privacy guarantees.
* Non-targetability: An attacker should not be able to attempt to compromise personal data that belongs to specific, targeted PCC users without attempting a broad compromise of the entire PCC system.
* Verifiable transparency: Security researchers need to be able to verify, with a high degree of confidence, that our privacy and security guarantees for PCC match our public promises.
- - - -
Second, according to their press release ([1] and a sibling comment elsewhere in this chain), they’ve been trying to find a way to allow interoperability without giving full access to everything. Unsuccessfully, so far. So it’ll be interesting to see where it goes, but I’m sympathetic to their current stance.
Could the restriction not be the device owner choosing to use it? If some rando vibe coded an app and the os told me all the things it can access, I'd probably want to trust the developer before installing it. Why do I need to beg Apple's permission to use software better than their first party offering?
Because you made the choice to trust Apple when you bought an iPhone. And while you may make a deep study of who is providing your alternative AI app (is that even possible with openAI or Copilot or Gemini?), the average use will pick something shiny and lose their savings when it transfers their bank balance outside the country.
> the average use will pick something shiny and lose their savings when it transfers their bank balance outside the country.
Couldn't you make a more believable straw man, please? The "Nigerian prince wants to send you billions" is really tired. Try something more emotional! Hackers will steal your kid's photos and post them on pedophile forums or something. This will resonate better with uninitiated and allow to easier lobby monopolistic practices. Good luck!
Just because I bought an Apple product doesn't mean I made the choice to trust them globally across everything I do on my device, when did this become a binary that the hardware vendor must also be the only trusted software and service vendor? I like my MacBook because I trusted Apple to build great hardware, a pretty okay os, and services I don't give a shit about. I won't buy an iPhone because Apple has removed the ability to distinguish between those things on that platform.
Surely there's something better we can do than say "the average user is a dumbfuck better consolidate all control with Apple".
I’m sorry, but the DMA is mandating Cambridge Analytica-type access to data. That whole scandal was people on Facebook granting a third party access to all the data they had access to. And Cambridge Analytica lied about how they were going to use it.
Facebook got roasted for this, but now the EU wants the same open data policy from every big tech company.
Oh wow... Cambridge Analitica had access to all the data Facebook had on you without you knowing it and without you knowing the extent of the data Facebook had on you. EU wants you to be able to knowingly install apps on your phone and give them access to the data on your phone you chose to.
China will get there first why because they’re run by engineers. Yes, they will have some stipulations but if you show them a good idea, it’s a good idea. They won’t stand on ceremony and say not invented here.
correct, but honestly it's much more interesting to have the new Siri on my phone than my mac, at least in my case. Very disappointing, having bought an Iphone 16 pro mainly for the new AI features which almost 2 years later are still not there to be used.
Both the EU and China put enormous risk on misstepping. If you raise the cost of distribution, even just through time-cost, don’t be surprised when it fucking costs
The OS restriction aren’t merely based on the region settings, they are also based on Apple Account region/country and on the detected physical location of the device.
Most of them are, some are annoyingly missing. It’s possible to install apps from two different app store accounts, but it’s 10 times more annoying than on Android. Additionally, there are some EU only features, such as third-party NFC payments.
Apple’s performative DMA outrage is getting more pathetic by the iOS version.
You can easily change your account region back and forth on iOS. Meanwhile Android has a bunch of extra checks and includes a 12 months delay before being able to switch it again.
A family member lost years of irreplaceable photos and paid apps due to account lock, and doing this across two countries they were legitimately residing in triggered it.
In my experience it only locked me out (of just the payment or account settings I think) for like 24 hours when I switched like 4 times in a single day.
From my personal experience I found it easier as a user. I had no balance or subscriptions at the time so I could quickly switch regions back and forth to get all the apps I wanted.
And there was a lawsuit and those users will be compensated.
This new update for 17pro is no longer misleading
“Apple has agreed to pay some iPhone buyers a collective $250m (£184m) to end a lawsuit accusing the company of misleading people about new artificial intelligence (AI) features and capabilities.”
What Apple said was that on-device models would be from the iPhone Air/17 Pro models and newer. That doesn't mean that the iPhone 15 Pro to 16 wouldn't get Apple Intelligence via Private Cloud Compute, though that probably means they won't have on-device models. I don't remember the marketing for those phones promising on-device models, just Apple Intelligence capabilities in general.
I have an iPhone 16 and I don't feel like I'm getting the shaft. It sounds like there's one specific customization thing that'll only work on the pro models? Not an issue. At some point in a few years I'll buy a new phone and it'll be the latest again, and then it won't be.
I'll probably get downvoted, but my guess is that they realized they need a cutting-edge local model that would only work decently on the most powerful iPhones. And I see why Apple would choose to give no AI at all rather than use a weaker model that will hinder user experience.
I strongly believe Apple can win the consumer AI space. They have incredible distribution and hardware. They just haven’t executed at the application layer yet.
what's worrisome is that they continue to fail at it. it's one thing to say "we're still hashing things out". it's another to parade around Image Generation features that are obviously widely not-cared-for and oftentimes actively disliked
Apple cares greatly about their brand yet this has hurt their brand like nothing else in the past decade
It’s worrisome, but it’s also worth pointing out that no one else has succeeded at it, or even gotten close. It’s easy to forget that there are not good well-integrated person AI solutions yet, it’s just chat bots. I think it’s just harder than a lot of people think it is.
but Apple isn't known to make grand promises and then not keep them, is it..? usually they just deliver what they say they will
yet i've been reading about "well they promised AI Siri two years ago and Siri still can't set an alarm right" in every thread even remotely related to the topic
i don't remember reading this much about anything else. it seems to have soured people quite a bit, at least in my internet bubble
on the other hand failing at it or pushing it to the edges until they figure out where if anywhere it would actually make sense (which is I believe the entire crux of their strategy) might equally reasonably be helping their brand.
I don't see strong evidence the average consumer is demanding 'AI features' in everything. I mean even amongst the technically inclined this is often bemoaned, anecdotally.
thats exactly what i'm saying they should be doing, but aren't. we're in agreement!!
why don't they just wait and not ship any AI junk at all? instead of promising a Siri AI rework, which then doesn't deliver? or Image Generation stuff that feels wildly put of character and generates tasteless and often downright creepy images?
not to mention that all of the new AI stuff they announced won't go live in China and the EU for a while.
why not do exactly what you proposed and wait it out? instead they seem to be trying to deliver AI stuff and just unable to.
there's also reports that apple execs held a secret emergency "oh shit what do we do about AI" type meeting.
Image generation seems like table stakes though, i.e. if you're going to offer a whole ass AI integrated with your operating system, it'd seem like a weird omission not to have image generation included.
Probably. It just feels so out of character. Especially the fact that it can't be safeguarded and generate sexual innuendos, for example, which Apple usually hates.
I'm not even entirely against genmoji. It's just odd for Apple to do.
If anything, shipping and sitting on the previous incarnation of Siri as a basic feature set to shut up iPhone users for a few years while everything else matured might be viewed as a very shrewd move ten years from now.
That's one way of looking at it. Alternatively; Apple had their pick of the litter at TSMC for half a decade, and Nvidia beat them to 5 trillion total valuation with their design chops alone.
There comes a point at which these execs cannot have more yachts. Apple still wields far more societal influence than Nvidia does, and will for a long time to come; and of course there's nothing stopping them from forming a partnership and putting Nvidia chips into an Apple device if it came down to it.
> it's another to parade around Image Generation features that are obviously widely not-cared-for and oftentimes actively disliked
I mean, that was one of many things, and I'd argue the least interesting by far. If the Safari extension creation thing is decent at all, that's a seriously cool addition. There's some real value shown in this most recent WWDC. Whether they actually release it this time is another question...
I mean have they, they have an outright majority share in iPhones in some markets, including the US, and lots of other stuff that sells reliably. Granted, I'm sure they'd love to have another blockbuster product, but having what amounts to "utility" status for a $1000 device isn't too bad.
ironically they respect user privacy the most and collect the least amount of data that’s why they lag behind ai i don’t understand why you would expect them to win
I just can’t imagine another computing device from a competitor company that 1) threatens the smartphone 2) apple couldn’t copy and mass produce at higher quality
If there’s truly an existential threat to its device business, Copy Well
The company that will challenge Apple will be one that possesses the capability to develop both an operating system and hardware at a high level right now, who is this company? There is one company that may one day, but that company isn’t in the west and no, it isn’t Nvidia.
Apple has built the strongest ecosystem of AI-capable consumer products in the industry. I suspect they deliberately chose not to compete in the AI data center race. Entering that arena would have cannibalized sales of their own high-margin devices, while further straining their already tight supply chain for memory and M-series processors.
Apple would never willingly pay Nvidia for GPUs anyway.
Why absorb supply chain pricing pressures and volatility when you can pass those costs directly to the consumer?
What’s real world test phoning home to the server if Google can make a Pixel look good by phoning home what do you think it’s gonna be different with an iPhone that six years ahead in the processor area? When it phones home?
Most people don't care that Siri doesn't do very much; if anything, people outside of HN are already sorta sick of AI features being shoved in their face.
Agree 100% ... but it would be nice to have some softer edges around those things, where I don't have to be so prescriptive and robotic in the way I make such requests. For example if I set a 15 minute timer, I'd like to be able to follow up five minutes later with "actually that 15 minute timer should have been 16 minutes" and not second guess whether it will do the correct thing.
I don't get it, this looks like what I'd expect it to show me if I asked it to find a certain message. I assume tapping on that message would bring me straight to that point in the conversation.
My complaint is this image is supposed to tell the story of how this stuff is useful or new. Finding a message from earlier in the same day and then paraphrasing the message instead of just showing it isn’t useful or new. A simple keyword search uses infinitely fewer GPUs and finds me the info as fast without the risk of hallucinations or datacenter driven environmental collapse.
But for whatever reason - because this stuff is mostly useless, because they’ve all got ai psychosis, or because of general dysfunction in the corporate structure - they can’t come up with a better example.
> But for whatever reason - because this stuff is mostly useless, because they’ve all got ai psychosis, or because of general dysfunction in the corporate structure - they can’t come up with a better example.
More likely, they just wanted an image that shows it can search your text messages for context. Most people still aren't familiar with AI, so having a very simple example image[¹] for people who may not have used an AI search tool before is useful.
[¹] I'm sure you and I can both think of more complex questions to pose to the Skinner box, but we're both familiar with AI and know what to expect.
I don’t think that’s true. Who are these blissfully unaware masses, and how do I join them?
I guess it depends on your definition of “familiar”, but it’s in the news every single day, and you can’t even google something without being slapped in the face with it. I feel like the only people left unfamiliar are people who have no access to the internet - and by definition they aren’t buying iPhones
I believe that is showing the results of a search, as in something like "What Plant was Aga telling me about the other day?". The website doesn't make that clear but they asked questions like that in the keynote (not that specific one).
Lol. That makes it almost make sense. It’s still showing me a summary of a text that uses exactly the same amount of space to convey exactly the same amount of information. Like, that’s just search - i’ve been able to get this info by typing “plant” into the search bar since they added spotlight in iPhoneOS 3.0 back in 2009
It’s also a text from 9:14am the same day.
If this is the second slide in your marketing slideshow, you clearly have nothing better to show.
The reason it's showing you what's in the text is to show you a search result. The text is essentially the proof. When the user asks what plant she was telling you about, they don't have the text in front of them. It could be 200 texts back.
Ok? But if I search for something just show me the thing I searched for (the proof). There is literally no need to repeat it in a slightly different tense. Who is that helping?
If I asked you over voice what Mike said he wanted for his birthday, it's usually less confusing if you just answer the question ("he wants snow boots") rather than started reading out words from another person ("I love it here at the snow, ...")
If that’s the story you’re trying to tell, where is the query in the image? The image alone doesn’t tell that story.
But also:
a) do you trust the llm to get it right 100% of the time? Because i’m gonna always read the original message to make sure.
And b) just excerpt the message “mike says i love it here at the snow … but i need snowboots” if you’re so desperate to shoehorn LLMs into everything, that’s just as easy a task for them as summarizing is.
The query isn't in these screenshots you're seeing online because it was in the video. This particular one I think was missing the query but the others like it had a presenter talking through what they were doing. Complaining about an image pulled off the internet out of context… Come on.
It’s not “pulled off the internet with no context”. It’s the second slide in their marketing presentation about what’s new in siri. The context they chose for it is all around it.
They’re adding vibecoded shortcuts (the high level scripting for Apple devices). Hopefully that means they worked out some of the long-existing bugs and missing features, but I’m not optimistic. Still, could be a useful tool, especially for less tech-literate people.
If they are ok with shortcuts being vibecoded, maybe it's time to expose a proper programming language to the end users as well.
All my automation shortcuts can be easily explained in pseudo code under 5 minutes, but it took me ages to put them together because that weird UI/UX forcing me to drag-and-drop squares around to manipulate data structures. Programmers hate it, non-programmers can't understand it, it is not designed for anybody.
It's funny how they went from AppleScript -> "oh writing scripts is too complicated for users, let's create Automator" -> "Automator is too simple, let's create Shortcuts" -> now with shortcuts being generated by language models they need a structured scripting language again.
AppleScript even had "dictionaries" declaring their commands and everything, would have been perfect way teach LLMs how to automate applications.
I've never been more frustrated programming something than when I tried to build a non-trivial Shortcut. Things that I could have expressed in a quick script took me literal hours to compose and debug using the WYSIWYG interface. And since there's no version control, any mistake runs the risk of dislocating an element or messing up an input/output connection and breaking everything permanently.
It increases the value app developers might get out of offering shortcut actions - similar to how the advent of MCPs seems to have kicked a bunch of SaaS vendors into offering a clean API, the advent of Siri being able to tap into shortcut actions - and script them - might make it feel more worthwhile to app devs to open up deep functions.
I just tested this myself. I wrote “flip the reduce white point toggle accessibility option in the settings app” and it worked perfectly. Run once to set it and run again to disable it.
I switched from android to iOS 7 years ago and I’ve actually been debating going back just because of how bad iOS is at AI. And every other facet of my life, I am finding ways for it to save my time on an almost daily basis. And yet on iOS, just finding something from a text Message is still a nightmare.
Siri seems to rarely get better and sometimes actually get worse.
Same problems on Android. Google Assistant was better in 2016 than it is in 2026, and the Gemini assistant is a joke. They can’t even integrate with their own music app (YouTube Music) properly.
I switched to iOS this year and I’ve been learning that the grass is not much greener. I do miss uBlock Origin. Maybe my next stop is GrapheneOS or a similar degoogled ROM…
There is a uBlock Origin Lite extension for Safari available in the App Store. Technically not as powerful (hence the “lite” in the name) but it’s perfectly blocking all ads for my usage.
It’s the keyboard for me. I don’t remember making as many typing mistakes when I was on an Android phone. It’s also infuriating trying to “select all” for copy or move the cursor to beginning with the spacebar motion.
I swear it’s like I’m using a phone in a dream sometimes.
Pixels have had it for years but I think it's in every android phone now; from the app switcher you can press and hold and the system will OCR the text and allow you to copy it.
Because so many things _still_ don't make that easy to do on mobile.
I don't think iOS has it from the app switcher, but pretty easy to take a screenshot and immediately select from it. (Or selecting from older screenshots or photos you open.)
Yes some of the best "AI" features Apple has implemented are related to picture search : you can describe what you want to find and it will consistently find it, even inside of videos with highlights or the moments you're looking for. Much less supposedly sexy than the SIRI AI / Apple Intelligence hype, but also much more useful
AI has just made it worse, both directly (search doesn’t function like standard search anymore but also nowhere near ChatGPT) and by comparison.
I can go ask Gemini questions that require it to get information from several emails at once like “which x vendor had the lowest price”. Im assuming it can do the same with my texts or, if not yet, it will soon. I had zero such faith with Apple.
I will wait until the fall and see if this looks like the germ of something actually useful before deciding if I’m going to switch.
Nope, it got markedly worse since AI. You used to be able to search for string literals, so if you remembered one obtuse phrase from a group chat, you could pull it up instantly. Now this 'intent' search will try to search for what it thinks you want, not what you typed. AFAIK, there is no way to search for literals anymore, thanks to AI.
Apple Passwords reliably updates passwords in its database before the password is confirmed to be actually changed. I've been locked out of accounts many times to this. They really need to focus on these basic UX issues.
All that data is lost when you migrate accounts though. I went from an old to a new 1P account and did the official way to copy (NOT exporting it to a text file and re-importing that way, actually copying it from the interface) and no version history persisted :/
Huh, yeah, I’m not seeing it there either. The macOS app is what I checked previously. iOS-only users might be able to see it at 1password.com. Weird inconsistency.
Still does not. My approach is to keep the file in OneDrive. On windows / mac it's just a file, on android it's via custom onedrive protocol handler but also seamless.
Yep. I get anxious when Safari starts to offer a new password for an existing account. Having access to previous passwords would be such great UX, but no, no such thing.
There's a 0% chance it will work. Most websites I've seen have one or all of:
* Force you to use email or SMS as a "second factor" to unlock changing password even if you know the old password
* A stupid idea of password complexity usually requiring one of a finite set of 5-8 "special characters" which is often only revealed after you've chosen a password that doesn't have them. Or in some cases even banning characters other than the ones they check for. There's a standard for this where you put a regex on the password field, which a good password manager will always use, but the kind of idiots who think limiting the entropy of passwords to increase security is the correct way to do things almost NEVER implement this.
* A maximum password length, even as short as 16 characters in many cases
* CAPTCHA etc.
Any effort spent on this would be better spent elsewhere, including even educating other companies on how passkeys should be used.
> Force you to use email or SMS as a "second factor" to unlock changing password even if you know the old password
Apple has detectors for codes sent via email or SMS, if your email account is one that is configured with the OS mail client.
> A stupid idea of password complexity usually requiring one of a finite set of 5-8 "special characters" which is often only revealed after you've chosen a password that doesn't have them. Or in some cases even banning characters other than the ones they check for. There's a standard for this where you put a regex on the password field, which a good password manager will always use, but the kind of idiots who think limiting the entropy of passwords to increase security is the correct way to do things almost NEVER implement this.
An AI agent can read the failure message and craft a new password
> A maximum password length, even as short as 16 characters in many cases
Same deal
> CAPTCHA etc.
While there's always the complex solution of scanning the image and trying to detect what is going on or slide the puzzle with enough of a curve to act like the motion of a human limb, there's also Private Access Tokens, supported by both Cloudflare and Google-provided captcha systems now IIRC. The OS uses an anonymous system to assert a single bit that there's proper browser chain-of-custody.
> Any effort spent on this would be better spent elsewhere, including even educating other companies on how passkeys should be used.
There are proposals as well to provide API to do upgrades from passwords to passkeys as well automatically. Nobody said the feature has to always use AI - but it may help the feature be robust enough for people to seek it out and try it.
Don’t forget those sites/apps that split the sign in process across five screens for bow good reason or those with mislabeled fields that password managers can’t pick up on.
I don’t think I’ve seen a single category of UX fail as hard and as often as auth screens do. It’s like at some point after 2015-2017 developers were struck with mass amnesia and forgot how to build decent login UIs.
Limiting the character set is done to reduce the frequency of "can't enter my password" support calls, not to increase security directly. Same with the maximum password length.
I don't think firms like the electric company or (payroll company) ADP are worried that I'll churn.
Also, the Venn diagram of "memorable" and "reasonably secure" really only intersects in the region of "Correct horse battery staple" phrases -- and the problematic sites I'm talking about nearly always limit length, which thwarts that type of password terribly. What is the purpose of maxlength on a password?? These shouldn't be stored in any form other than a hash, so unless long enough to pose a DoS threat during the hashing process, length is truly none of their business.
1Password has been able to do this for five+ years. Frankly, it doesn't even really need agentic AI, although a talented team could probably make it perform better with agentic AI.
I don't really believe in Apple being that quality team.
I mean every api/app/website has a different way to do this. If there was a standardized api that everyone could conform to to allow this automation I would be all for it. I assume 1p does this by writing custom code/rules for dealing with the most popular sites out there and then erroring out for anything else.
AI could potentially help solve those unpopular site/app/whatever edgecase.
Haven't you heard? Prompt engineering is dead. The cool kids are making Claude prompt itself. They're writing loops, not prompts. It's all about optimal tip-to-tip efficiency now.
“I don’t prompt Claude anymore. I have loops running that prompt Claude and figuring out what to do. My job is to write loops”.—Boris Cherny
heh I merely said prompt engineering because their efforts amounted to writing a prompt and sending it off to a model somebody else created to create some awful images
In Feb this year, when I analyzed KuzuDB's source code, I predicted Apple's reasoning to buy them was to introduce Siri with cross-app personal context. "..WWDC 2026 or 2027 introduces any “contextual intelligence” features in Siri that require cross-app relationship reasoning." https://medium.com/data-science-collective/i-analyzed-163k-l.... There is no confirmation on which tech is being used to achive that though
Google is a great developer, but a really shitty consumer company.
Nest devices are garbage, it has been like a decade since their phones were competitive with other vendors, ChromeOS was barely hanging outside of education center and now it's a zombie walking since Apple released their cheap Mac Nano, Gemini is a joke as a product compared to Anthropic and ChatGPT. The only things worth something are Chrome, Gmail, Workspace and their Cloud.
Google's power is in controlling your interface. How you interact with the internet, and now, how you interact with AI, is dictated by them. There is massive power in being the interface between you and information.
Not directly, but they control the most popular portal to the open Internet with chromium to shape ad delivery, and (depending on how dramatic you want to be) the shape of reality presented to their users based on their AI models. Not directly profitable, but damn that's a scary level of power.
No, I don't want Apple to read all my emails, text messages and photos. I'm probably not going to buy Apple devices any more if it really works like this, by default sending all your data into their cloud.
If I want to use AI, I want to be able to select the exact messages / photos which I want to send to it. Otherwise I expect the device to keep the data protected. I don't need any of these features either; I can remember if someone sent me a cookie recipe.
You might want to so some research. It’s the only company with a level of privacy on this.
Most of it is on-device. Private cloud compute does not store anything.
I dunno man, you could criticize 360 aspects of this presentation but you pick privacy, really?
It can make on-device tool calls using your data (web search, database search, email, app APIs) which are not private. PCC doesn't protect against being dumb with your data.
E.g. "the user asks if their Bitcoin private key is unique, let's make a web search".
Nah, they actually have everything open for auditing and you can do that yourself. They heavily advertise that and had been for a while since they announced that split model 2 years ago
This looks pretty promising to me. It will likely replace the need to set up OpenClaw for average personal users. The work of getting email, messages, and all the personal data on the phone as context seamlessly is not as straightforward as one might think.
I'm curious how the pricing will work. Would it be free up to some limit and then some subscription pricing? I can't imagine it can be free unlimited usage given the price of serving these models.
Craig mentioned near the end of the keynote that compute intensive things (like image generation) will have rate limits that can be increased bundled with their iCloud + plans. I imagine any request that gets routed to their cloud compute will be subject to limits as well. He positioned it as a value-add to their existing subscription but I suppose that can change.
This is why OpenAI thinks it needs to build its own physical devices. If Apple is only allowing its own AI to operate at the OS level, then that leaves OpenAI with no choice but to build their own.
Meta also realized this and attempted multiple times to build their own hardware but they've given up each time. They started as early as a partnering with HTC in 2011 to make a Facebook phone.
Quite frankly, I'm kind of excited to see what OpenAI can build. I think an AI-first phone could challenge iOS and Android. It's a new paradigm and if OpenAI gets it right, it'll be very hard for Apple and Google to pivot.
I personally think chat + code is the future of apps. For example, I find myself wanting to do many things inside ChatGPT instead of traditional app because I can tell it to do things that are simply impossible on a static app UI. For example, I have some data I want to send to an app but before I do, I want ChatGPT to clean the data in some way first. And then after the data is uploaded, I want ChatGPT to pull some data off the API and make charts that I want to see.
I imagine a world where very intelligent models run at 10k tokens/s, app building is extremely standardized, and it simply builds any app you want inside the OS. IE, if you want a dashboard of your health data, you ask it to build it almost instantly exactly how you want it. I'm already doing something similar today but it's slow and not easy to do for non-engineers.
> If Apple is only allowing its own AI to operate at the OS level
Incidentally, that’s what’s preventing Apple from rolling out their OS-privileged AI in the EU, as the EU mandates equal access for competing AI products. It will be interesting how this plays out.
> Apple would have to give any virtual assistant direct access to users’ private data — and the ability to directly control other installed applications — as soon as Siri AI is made available in the EU
> do they have to let Anthropic and Deepseek run their models on it too?
provided it gets big enough, yes. the EU's position roughly is "if this hurts an entire market just to benefit you, and lots of people use / rely on it, then you gotta allow it"
Then the EU’s not gonna get anywhere the US and China will just sail along. The lawyers in the EU are essentially cutting their throats building hardware and operating systems cost real money and time which is one of the reasons Apple is using Google’s model for $1 billion a year. (when Apple finishes their models in house, Google probably will be dropped). Like dropping Broadcom, Intel, or Qualcomm in 2027-28).
It takes real time to drag along five ecosystems. That is the main reason it’s taking Apple longer than they’re so-called competition? Noticed that Google and Microsoft only do bits and pieces. Microsoft has no mobile and Google at its heart is an ad company the processor is six years behind.
that's the most common criticism. can't say i disagree.
though monopolistic practices are unequivocally bad and used to get struck down. one may argue this is just another instance of disallowing monopolistic behavior.
the DOJ used to have sharper fangs than what the EU is doing now
Why can't Anthropic or Deepseek take the big risk to develop their own phone? It doesn't seem right that they can simply use EU laws to hop on the ride for free without taking the same risks.
As a consumer it also doesn't seem right that Apple can just use all their private APIs that no other company is allowed to use to tell me what I can and can't use on my phone. If I want Anthropic to have the same level of access that Apple grants itself on _my_ phone, I should be able to do so.
As if Android and Windows don't also use the same tactics. Windows used to not let you uninstall Edge and Android used to just default to Chrome and Google Search, now they have to prompt you what you actually want to use. If the big companies are left unchecked, of course they'll use their size to bully other players to increase their own profit, but always at the cost of the customer.
Do you rather want your refrigerator, washing machine and coffee maker send all the data it collects about you during your usage to some company you don't even know? How is it ridiculous to want to be able to make your devices do what you want to do, not what the big company behind it wants?
1. A lot of those more recent machines do actually connect to the internet.
2. Since you want to run your own software on them, but you can't buy open hardware like that, did you build it yourself then as you suggested? "If they don't exist, make them yourself."
the EU feels like this falls under monopolistic practices, which it has deemed illegal. the buck stops with the politicians - there's no reason other than "the EU thinks it's bad for the economy and should thus not be allowed"
OpenAI is not a hardware company or an OS company and they hardly have the money to pay the real cost in the long-term, if Amazon, Meta, Google and the Linux crowd are having trouble, OpenAI is not gonna get the job done.
> Quite frankly, I'm kind of excited to see what OpenAI can build.
I’d go further and claim if controlling the hardware is a existential threat for these companies, their ability to deliver or not should be seen as an indictment of the entire “llm formulation of ai” era we’re currently in. If llms have the potential they claim, then empowering them to design a best of breed phone should be table stakes.
>Meta also realized this and attempted multiple times to build their own hardware but they've given up each time. They started as early as a partnering with HTC in 2011 to make a Facebook phone.
I was working in cellphone sales at the time and I can tell you no one wanted that phone back then even when Facebook was massive. An easy to hit facebook button was not a value add anyone was begging to exist.
Although with how many phones now have stock forced installs of Meta apps perhaps they won their con in the long game.
Similarly no one really wants a physical AI device, and attempts at such are pure techbro hubris on the companies part.
I want an AI assistant that I can use truly hands free. I keep my phone in my jacket when I'm riding my motorcycle. I want to be able to start, stop, adjust, and check details in route guidance. I want to be able to ask what the weather is like ahead on my route. I want to be able to ask it to start looking for a sensible place for me to stop for fuel and/or food without making me do a big detour.
Actually I would also quite like better driving directions, since I can't look at the directions on a screen.
They need software companies to give a damn and improve apple maps/gmaps/etc.
Or perhaps pull over and adjust your routing like most of us do, or maybe bolt a sacrificial phone with maps on it to a holder. If you don't want to risk your main phone.
Todays pulling over to adjust planning is yesterdays pulling over to make a call, and yesteryears pulling over to throw a map on the hood.
Riding on a motorcycle is already dangerous enough with the average land tank driver on their phone. Talking to your AI assistant while riding at speed sounds like pending split focus disasters waiting to happen.
The chatbots(ChatGPT, Claude et al) showed Apple exactly what can be done, the user base is already well primed. So this is a product definition done for them to execute. If done well they will be able to provide a much stronger integration into the day to day use cases than the chatbots, and can siphon off user time from them. This time around the end to end is easier with Apple Intelligence and more importantly llms doing the work Apple is floundering at. So I am hopeful, but I still see the os/app level integration as not enough in terms of functionality to make it a hit. The primary use case for llms is still conversations and search. Apple should be focusing on that aspect primarily and also add the os/app level integration as a bonus - as something only they can do. If they just do the latter, it will not be as much of a success. Let’s see how they execute.
EDIT: To provide meaningful chat functionality they have to either eat up the cost or charge a subscription for it. This will be first time they charge for Siri - a product that doesn’t garner any positive reviews. This gets even more interesting to watch
> This will be first time they charge for Siri - a product that doesn’t garner any positive reviews.
It seems like revisionist history to say that; lots of people were sold on iPhones years ago because of Siri. They have one of the few business cases for voice assistants, which are notoriously difficult to actually monetize, that actually makes any sense, since "selling iPhones" is meaningful and "selling a subscription" would be nice on top of that.
That's not what that is showing, it isn't reading a text you just got. It's you asking about the plant she told you about, it finding that text message, summarizing, and also proving with the text that it's accurate.
> Schiendelman sent you a message on hacker news clarifying what its showing. It's you asking about the plant she told you about, it finding that text message, summarizing, and also proving with the text that it's accurate.
I've not seen anyone discuss yet what kind of impact these AI features are likely to have on battery life. If they are running small on-device models, this will have a measurable impact on battery life with regular usage.
Before they add AI they better fix the frigging search function in settings, it is horrible, you need to know their exact words, and Apple has a funny naming sense. Hierarchies nested so deep you never find anything. I come to use Claude or ChatGPT to tell me the right incantations to find a setting.
I don't even know that it's knowing the exact words, it just seems like sometimes the search decides it's just gonna give you no results because the vibes were off.
I'm almost sure that sometimes searching the same thing will give you the result and sometimes it won't.
I wonder how much of Siri AI is Apple-developed and how much of it is Google-developed as a result of Gemini. The a) search demos and b) image generation demos seem unlikely to have been done by Apple alone, the demos being closer to Google Search and Nano Banana respectively.
It looks almost entirely like gemini. The images they showed are obviously nano banana, and the text responses are almost obviously Gemini (I say as a somewhat frequent Gemini user).
I'm sure they customized some of it, but this looks basically like Gemini integrated with iCloud instead of Google Workspace.
The images were the biggest tell, generating using a reference photo of a person, at least Gemini and ChatGPT have two distinct styles. ChatGPT is a little less uncanny valley than Gemini which tries to be too realistic looking, in a bad way because it tries to preserve the original person in the photo, but still can't seem to help altering facial features.
The text responses had Gemini's verbosity. Asking ChatGPT to show me iconic dishes from both Brazil and Morocco (Apple's example), is much cleaner, less verbose. Quick list of dishes and links to the recipe. Gemini just spews a wall of text and bullet points and goes on and on with fluff. Tons of "What this dish is" "Why it works" Same with its frequent use of tables, which I see less of with ChatGPT.
Each Siri demo they did in the keynote had that hallmark verbosity I typically get with Gemini without prompting it to not do that.
The only thing you saw was phoning home, that’s what’s gonna be interesting when Apple releases their version. Is it phoning home 100% of the time or can you turn off the Internet and have it perform in the same way, there will be plenty of YouTubers that will give it the test a test I might add that they haven’t done up until this point for anything that Google has put out?
Is Siri any more or less than “just” an agentic harness such as OpenClaw? How much of what that harness does is up to the LLM or the harness itself?
In my mind the Gemini LLM defines the bounds of capability and capacity, but any actual functionality or usefulness (or lack of) comes from Apple’s Siri harness.
I was wondering the same. I have to imagine it's mostly Gemini, unless Apple has a big, secret, SotA foundation model no one has heard of? But if it is Gemini, how does that work with their Private Cloud thing? Are they able to load the Gemini weights into it?
It’s almost all Gemini and the “Apple local models” part seems to just be image embeddings/descriptions powering new spotlight and the like which is also likely someone else’s model.
This sounds like a take that you'd probably only find on HN though. It seems dubious that regular Apple users see any issues with the software stack or with Siri, certainly not to the point that it'd be a branding problem for Apple.
Next time you're speaking to a close non-techie friend or family member feel free to ask them how often/whether they ever use Siri, and then ask them why not, and you'll realise that no, this isn't some HN-specific peculiarity.
Quite possibly one of the most lackluster WWDC's I can remember. And I've been a fan of Apple/Mac since the late 80s.
Hmph.
That said, I'm THRILLED they claimed to "fix" the border radius snafu of Tahoe. Go ahead and push that now with the next Security fix. We won't mind at all.
I placed eggs into boiling water and because my hands were wet I used voice and said "hey Siri, start a timer". It replied "you'll have to unlock your iPhone first", so I hovered my face over the bench the phone was sitting on, then the screen rattled and prompted for the numeric passcode. Ugh..
Using Siri essentially required me to use my hands anyway, so what's the point of voice?
I'd very seriously consider moving away from iPhone to a device that treats voice AI as a first class citizen (presently I mapped the 'double back tap' to open grok voice chat, and triple back tap to end it, which is a wonderful improvement over not having these, as you can do that pretty easily, even while driving etc).
They do (one hand), but it doesn't require anything precise (like finding the app and clicking on it, which you can't really do while driving). If Siri could open an AI voice chat, that would awesome. But I'm not counting on that within this decade.
The only way Siri will meaningfully save my time would be if it could doomscroll for me. All of the things Siri claims to automate here would save me maybe a couple of minutes total over a week, wouldn't make the tiniest dent in my screen time.
I don't understand. Are you so little in control of your own self that you now need an embedded AI to do something for you that was already a complete waste of time and energy? Just delete the apps.
The point I was trying to make was that consumer AI on mobile devices won't save people time in any significant way because people don't spend a lot of time on their phone doing things that are automatable. Everything mentioned here is a tiny portion of people's phone use compared to things like doomscrolling, video calls etc. So there is no real value to any of these tools.
What is not demoed are basic AI capabilities I need today to effeciently use my iPhone. Examples:
- Siri, send the latest 20VC podcast to my watch. I'm going jogging.
- Siri, do not start Apple Music. I do not use that app. Start Podcast.
- Spell check and Auto-Correction on iMessage should support French and English. I switch from the 2 all the time.
- Etc...
This feels like it could be solved with a list of permissions that the user has to turn on when using 3rd party AI.
Apple already:
1) requires developers to submit ID to publish an app on the appstore (at least I had to after ~1000 downloads to be able to publish an update)
2) has strong kernel enforced memory integrity and disallowes arbitrary code execution (unless explicitely approved for games like roblox, jitting not allowed tho has to be interpreted).
3) reviews every app update.
I feel like this is nothing more than Apple being angry that they have to allow people to actually choose what AI they want on their phone. This is particulary interesting if anthropic and openai decided they want to add siri ai override to their apps allowing them to take advantage of the apple ecosystem without signing some kind of deal like they had to with Google. I assume behind closed doors Google had to make some sacrifices for them to be the model powering siri.
> I feel like this is nothing more than Apple being angry that they have to allow people to actually choose what AI they want on their phone.
It's really just Apple being angry about the EU's DMA endangering their golden goose (App Store revenue) and using any meaningful new functionality as a bargaining chip.
They've done staggered geo launches for other features in the past many times, both before and after the DMA was passed, and in this case there's even another great reason to not want to globally launch all at once (AI inference server capacity). If they can at the same time market it as part of their ridiculous turf war against the European Commission, I guess they just have to take the opportunity.
I struggle to believe there isn’t a significant fraction of EU citizens who are frustrated with the EU’s laws.
At the very least to the extent that the whole setup limits national sovereignty.
Your comment comes across as though you expect us to believe EU citizens are a homogenous whole, who happen to align with your perspective on this matter.
As someone who’s now switched over completely to USB-C and has saved money and reduced cable complexity incredibly, even if I ever get mad at the EU it will never be over Apple, which has time and again demonstrated itself to be an incredibly bad faith actor.
They complain about things their competitors are able to implement with no problems at all, and then once their tantrums are not sustainable anymore, they’re apparently able to solve all the problems as well.
For a company that sells itself as a design + engineering firm, they seem to have very little confidence in their ability to design or engineer their ways around constraints pretty much everyone else is able to.
On the contrary, it reads to me like they've simply been around, and this is the general impression they gathered. Which may still not even be true, but it makes a whole lot more sense context wise, and is pretty darn different to the conveniently malicious motivation you're proposing. And with this, now your own "royal we" is similarly rendered deceptive.
> At the very least to the extent that the whole setup limits national sovereignty.
That's how anything grouplike works indeed.
> I struggle to believe there isn’t a significant fraction of EU citizens who are frustrated with the EU’s laws.
Sounds like something that'd have polling data coverage?
We've banned this account for breaking the site guidelines and ignoring our request to stop.
If you don't want to be banned, you're welcome to email hn@ycombinator.com and give us reason to believe that you'll follow the rules in the future. They're here: https://news.ycombinator.com/newsguidelines.html.
I would call those people uninformed. It's completely reasonable for Apple to be hesitant to roll out any new integration features within their own ecosystem in the EU.
iPhone mirroring for example. Seems like practically 100% chance that if they put that out in the EU they'd be facing lawsuits for not making it work with every Android and Kindle and digital pregnancy test on the planet. And making it an open API right out of the gate is a much bigger undertaking than just making it work with your own devices through a proprietary API that you're free to break at any point and just update your devices to work accordingly.
Yes, it is harder to support APIs compared to… only having a completely closed system.
It would take a trillion dollar company to have the kind of resources that kind of thing would take.
The fact Apple thinks it should be exempt from rules that try to impose some fairness and choice, simply because “it’s hard” is the kind of thing that makes me feel they should be forced to be broken up in some fashion that limits the conflict of interest. They can be a petty platform owner who extracts rents from every developer OR they can offer things like apps and services on top of a platform. They are proving themselves too evil to be both.
> [...] for not making it work with every Android and Kindle and digital pregnancy test [...]
If only there were a way to design a system in a way that didn't even require you to know who's on the other side of an interface. Some kind of... protocol specification maybe.
Imagine the things we could do, like making a phone call to a different brand phone than yours, or a web server running a different OS than the client, maybe even one that the other side has never heard of!
> Given the serious risks to users, Apple designed a solution called Trusted System Agent — an intermediary that would allow virtual assistants to safely access the same features and capabilities as Siri AI for devices in the EU. Apple also shared a plan to launch Siri AI in the EU while gradually rolling out this new solution over an 18-month period. The European Commission said no. In fact, the European Commission did not agree to any of Apple’s proposals.
I'm extrapolating (there is less detail in that press release than I expected from your comment), but this sounds to be like it would be the thing that enables such a "list of permissions". I would be curious to know exactly what this agent entailed and why the EU did not approve it.
"Trusted System Agent" imo sounds like an apple approved agent which would only be available to companies that accept apples (likely unreasonable) demands and would completely lock smaller companies out of the ecosystem.
My limited understanding is that it would be a local model that exists only to determine a limited set of local information necessary to answer the user's request. This request and information would then be shared with the third party. Third parties would otherwise not have access into the local semantic model based on user personal data.
Does apple's own siri need to pass requests through their gatekeeper AI? I bet it doesn't. Personally, I'm generally happy with any answer apple comes up with so long as they're bound by the same set of restrictions as 3rd party companies. I feel like that's the only way to make sure apple won't "accidentally" hobble their competitors. (Like they did with their ridiculous 50c per app install fee for 3rd party app stores).
I want Claude and OpenAI to be able to compete with Siri on an equal footing. Just like Apple maps has to compete with google maps. Competition pushes companies to make better products.
Hm, I didn’t even consider that it could be an “agent” in the AI sense. I assumed this meant a service that runs on the device and interposes on requests to access privileged resources and enforces permissions checks on them. That is, the classical sense of the word agent in computing. Perhaps you’re right; in any case I don’t think there is really enough detail here to go off of.
It’s kind of funny that the EU’s regulation here would force Apple to allow options that are worse for user privacy. Apple is the least incentivized to farm data from its users; in fact, that’s a huge selling point. They mentioned it over and over and over in the WWDC keynote today.
In my opinion, Apple is doing the right thing for users. It’s not like they have a huge revenue stream here. Yes, there will be some features or usage that require iCloud plus or whatever to cover incremental cost, but I genuinely believe that they don’t want services creeping in that break their trust with users or their privacy-first reputation.
Apple’s decision (users will have a less powerful product because we’re not vacuuming up their data and using it for profit) is exactly the kind of thing the EU should want. No country has appropriate data privacy guidelines for AI (yet) so opening up choice can’t provide alternatives.
(To be clear, I’d be fine with Anthropic here, but am fine with this state. Maybe because I’m so used to Siri sucking that I’ve given up hope.)
The right thing for users would be to allow user choice, and for Apple to compete fairly.
Apple allowing third party access doesn't automatically mean user data gets hoovered up by OpenAI, Anthropic, etc. It just means users now get the choice, if they want to make that choice. Users could stay with Siri/Apple if they care about what Apple is offering, or choose to accept the risks and terms of service with other third parties.
The EU isn't saying "you must preinstall every competitors offering" its "you must offer the ability for others to hook into the same APIs to be able to offer their own assistant on par with the first party option."
The user still remains in control by virtue of their own choice.
I never understood how any regulatory body is going to decide which APIs in iOS must be made available to third-parties to hook into. So what if I'm a third-party maker of TCP/IP stack and I want Apple to offer me the ability to sell my custom TCP/IP stack to my iOS customers as a replacement for the stock TCP/IP stack that ships with iOS. Clearly no regulatory body has cared about that because it's too niche of a space?
So some government official will scour the entire API surface of iOS and decide which ones Apple needs to expose to third-parties? They have already decided App Store and Payments APIs need to be made available. Now it looks like they also expect off-device foundation models need to be made available to third-parties.
What about making Apple Watch specific APIs in iOS be made available to all third-party watch makers so any one can bring any smartwatch and use it just as effectively as the Apple Watch with an iPhone? What about all the AirPods specific APIs that lets Apple offer a better experience with AirPods than a generic bluetooth earbuds? What about Apple Pencil? And so on... If you go down this path, the list is endless.
> What about making Apple Watch specific APIs in iOS be made available to all third-party watch makers so any one can bring any smartwatch and use it just as effectively as the Apple Watch with an iPhone? What about all the AirPods specific APIs that lets Apple offer a better experience with AirPods than a generic bluetooth earbuds? What about Apple Pencil? And so on...
Don't threaten me with a good time? All of those seem like great policies. The fact that I cannot use an apple watch with an android phone is ridiculous, and vice versa as well.
Should you be able to use a Samsung SoC in an Apple phone?
At some point this is just a debate about vertical integration. Apple can deliver better experiences with it, but of course it limits user choice.
Many people want fully modular, open systems, which is lowest common denominator.
I can see both sides of the argument, but I am so skeptical of regulators deciding what can be integrated or not. If modularity is better for consumers, why don’t they prefer modular systems?
At the very least I think there should be a very clear tradeoff; right now the EU seems to think they can regulate their way to all of the benefits of vertical integration while outlawing vertical integration. I don’t see how anyone could look at that with a straight face.
> Should you be able to use a Samsung SoC in an Apple phone?
How did we go in less than two comments from providing access to APIs that are already present, implemented and actively used by Apple (who in their holy wisdom deem us mortals not worthy to access these the way we choose) to a completely different hypothetical of requiring actively building support for another companies hardware?
Such slippery slopes really aren't helpful, nor in any way comparable to what the DMA actually intends or states.
“Smartphone” has become a mandatory thing everyone is required to use to function in society without major friction.
Businesses hate supporting a ton of distinct platforms, as proved by the developer marketplace killing Windows Mobile through refusing to ship apps for it.
This suffocates any third entrants just like the FPTP voting system suffocates third political parties.
Yes I'd like some of these too but at the same time I get an uneasy feeling when I think that some potential idiot in a regulatory body in every country is now going to decide which API surface needs to be made available to third parties. If they take it too far, they could end up making nonsensical choices and kill innovation.
Good. They are making an operating system. User choice and competition matter. I know Apple would prefer to allocate more resources to Liquid Glass animations and burying more UI elements inside “…” menus, but I personally think I don’t need any more innovation above the OS level from Apple. Especially because 80% of their changes to the application layer in 10 years have just made their platforms worse.
Let them ship a stable platform that allows applications to do tons of useful things, even when you don’t accept a mega-package of apps and services all from the first-party vendor that locks you in.
If Apple built houses, you would have to jump through hoops every time you use a microwave or lamp you didn’t buy from Apple.
> What about making Apple Watch specific APIs in iOS be made available to all third-party watch makers so any one can bring any smartwatch and use it just as effectively as the Apple Watch with an iPhone? What about all the AirPods specific APIs that lets Apple offer a better experience with AirPods than a generic bluetooth earbuds?
That's...literally the point of these regulations. Sounds great to me.
Creating competition where it would not otherwise exist is the essential nature of the EU. Originally it was mostly about forcing protectionist member states to accept competition from other member states. But they extended the approach to breaking perceived natural monopolies a long time ago.
The exact rules ultimately don't matter, because the EU is after outcomes. If the current rules don't lead to the desired outcomes, they will keep changing the rules, until they get what they wanted. (Or until their goals change.)
Destroying competition by removing the consumer choice for vertical integration in service of strong security, privacy, reliability, etc.... is mistaken.
It's competing at the wrong level.
The iPhone is a toaster. Nobody's up in arms about whether the toaster takes other manufacturer's crumb tray. It's a television, and nobody's demanding QLED and OLED be swappable. It's a console. Xbox doesn't play PS5 games. It's fine.
There's no real line between hardware / firmware / software / malware ... For what Apple offers consumers, every layer of whateverware should be trusted.
Drawing imaginary lines based on the embodiment or substrates for logic gates is mistaken.
There are lots of phones. Lot's of different philosophies. Stop taking away consumer right to pick a philosophy and design for an end to end experience. It's fine.
Nothing about allowing others equal access to the OS means that someone can’t still choose Apple’s first party services and products.
It’s not an either/or thing, it’s about preventing so called gatekeepers from anticompetitive behavior via favoring their own accessories and services while simultaneously preventing any others from possibly competing.
There’s no valid reason at all a third party smartwatch shouldn’t be able to integrate to the same level as an Apple Watch. No reason third party Bluetooth earbuds shouldn’t be able use ADWL for automatic device switching, etc.
Want to still use only Apple? Great, nothing says you can’t. But at least it would be user choice and there would be actually competition which would lead to better products for all.
Can’t believe I lived to see the day that people on HN start defending vendor lock in and closed platforms as a good thing. Have all the hackers retired?
Oh, me, me! I spent a few years being responsible for a significant bit of DMA review and CYA and responses to regulators.
I’ve read all of it, multiple times, and been grilled by EU regulators (vicariously, via corporate lawyers).
It still boils down to general guidelines that it’s impossible to know if you’re violating before the fact, and they will not even approve/reject proposals in advance. It’s basically “go read the act yourself, and ship what you think is compliant, and you’ll know whether we interpret the words the same way by whether or not we fine you.”
> It still boils down to general guidelines that it’s impossible to know if you’re violating before the fact, and they will not even approve/reject proposals in advance. It’s basically “go read the act yourself, and ship what you think is compliant, and you’ll know whether we interpret the words the same way by whether or not we fine you.”
Companies want to know exactly where the line is so they can figure out how to comply with the letter of the law while doing as much as possible to get around the spirit of the law. This has been demonstrated over and over again. It isn't the job of the regulator to help companies with this process.
So you’d be cool with speed limit signs that said “hey, don’t go too fast” and no specific limits? And the cops decide who to pull over on vibes, reputation, mood?
I’m more of a rule of law person myself. If there’s a law that must not be broken, and breaking it results in penalties, it seems insane to me to not specify it in advance.
Sure, big tech is largely evil. Arrest ‘em, find them, IDGAF.
But pretending that DMA and related regulations provide enough information to ensure compliance is willfully ignorant. The regulations are designed to allow selective enforcemen.
>speed limit signs that said “hey, don’t go too fast”
Yeah we already have that. We have the words 'SLOW' on the road that ask you to slow down from the current limit for the hazard ahead, but pulling you up on this is officers discretion.
Bad example, as there is a posted speed limit above which you are positively breaking the law. Discretion may lower the limit.
The DMA doesn’t have the objective measures. It’s all discretion, all subjective, all post hoc.
Which, cool, some people like the idea that police target those people and need flexibility to make life harder for undesirables in ways they would never do to high status people. I don’t personally like that, and I don’t think tech regulation should work that way.
VPN is not a replacement TCP/IP stack. I literally meant the TCP/IP stack in the XNU kernel. It might be an esoteric example but it's not that far off. DMA already forced Apple to open up browser engine layer so third-parties can now bring in their own browser engines in the EU and are not restricted to using just WebKit.
True. Will add, device must be supervised to use VPN always-on which is possibly sensible albeit annoying (would have to reinstall iOS and set up as new I believe).
> Apple allowing third party access doesn't automatically mean user data gets hoovered up by OpenAI, Anthropic, etc. It just means users now get the choice, if they want to make that choice.
Apple is also restricted in the sort of consent prompts they give the user. That could matter when a non-technical users is prompted by a third party app to effectively allow unfettered access to all user personal data on the device.
Sometimes when you look at the functional requirements for a feature it turns out to be a bad idea. In the EU, functional requirements can come after-the-fact from regulator interpretation of the DMA. Until Apple determines what those requirements actually are going to be, releasing a potentially harmful feature is irresponsible.
And why is that a good thing? The average user can't even spell Anthropic. Why do you think they can safely pick a third-party model provider that could harvest the hell out of their conversations? The control of ecosystem is part of the privacy and security. My mom's Android phone has like 100 apps that she had no idea how they were downloaded. For real user choices, the vast majority of users just want a phone that they can trust and don't have to be a techie to avoid being exploited. They can choose to buy a phone that can be built from legos, OR they can choose to buy a phone from someone they trust to get the privacy and security taken care of for them. This is the real user choice.
There is a vast asymmetry in knowledge and capability to make those choices for most users. Most users will press agree and consent to things they don’t understand to get to the next screen, and while that would be terrible for the individuals I think it’s also important to look further at the ramifications because probably Apple would be blamed.
They can't under GDPR. The DMA is for market access - there are other laws for privacy. Those require use commensurate with what is needed for the service, so anyone who e.g. scraped all of a user's local info and stolen it would be breaking EU privacy laws themselves.
This is not complicated. Even in the US, every other industry is regulated to your benefit, you're just used to it and haven't realised. Digital technology obviously needs to be too. And yes, you have to do it properly.
Apple ad revenue is ~10% of rev, with Google deal, and growing. New management is going to turn it less privacy focused company, because Apple needs to pursue growth.
Apple may have a true argument that their version of “AI” is currently the least privacy-problematic, but it’s not a compelling one. First of all, it sets an awkward precedent. This special status is mostly based on vibes, and on how we know the other guys’ business models will push them towards more data collection. But doesn’t make sense to say we allow anticompetitive behavior depending on how nice to how “nice” the aspiring monopolist is to their users.
Google can easily argue that if Apple gets to rule over a walled garden, zero-API ecosystem where no one else can compete, then it’s right that they can too, regardless of how privacy-respecting they are or aren’t.
if for a second you believe that what apple says the regulators told them is the same thing as what the regulators told them, i have a cow farm under the titanic to sell you
This comment casts aspersions while making zero specific claims of wrongdoing. If you have something specific to say that goes beyond the vibes of "everything and everyone is corrupt and evil," that would at least be worth hearing.
Your premise is incorrect; if apple truly wants to do the 'right thing for its users', it would allow choice. The fact that the current crop of likely alternate choices include quite a few companies and offerings that seem far more user hostile than apple's offering doesn't change that fact (it merely raises separate concerns that there need to be more laws such as the EU's DMA, not fewer).
However, even if your premise is correct, it does not matter.
In the end, trying to manage such products (require massive investment, have network effects, offer significant gatekeeping and rentseeking opportunities) is extremely problematic.
On one hand, the market cannot do it properly: There are tons of externalities, and, like e.g. building out rail, the absolutely gigantic barriers to entering the market means the existing players merge into a monopoly or oligopoly.
On the other, the product is too complex and too dependent on continuous evolution to officially turn it into a state-controlled / state-run monopoly (the solution many countries have deployed to solve e.g. how rail, or medical insurance, or road networks, end up in a terrible state if left up to the market).
So what is one to do?
The current crop of mostly US led large companies seem to have gone with a 'just trust me, bro!' argument, with some 'AI is so important you cannot put up any roadblocks at all!' sprinkled in.
And yet these companies time and again prove that they can't be trusted. Which is obvious and logical: Companies must conform to the law, but are otherwise amoral. Or rather, their 'moral' compass has nothing to do with human moral compasses: They must earn money for their shareholders, in whatever legal way they can find that is most efficient, paying as much attention to future company growth and health as its shareholders desire. That isn't just 'what they are incentivized to do' - that is what they are legally *required* to do.
And yet you've gone with a motif of 'but apple is the one company that is doing it right so lets just trust them.. bro'.
There *is* a solution:
Use the fact that the state has powers of persuasion that companies simply do not have. The threat of law, and the monopoly on violence.
Essentially, a state can simply tell a company: The populace have spoken and they value X (say, privacy). They value it a lot. You will deliver. At low cost. This is not a request, it is a demand. If you don't want to or can't, then we shall write laws to regulate you and then *everybody loses*.
Conceptually this works, in a weird game of chicken / madman theory: If the corporation in question believes that society will regulate them into oblivion unless they comply with society's demands even if this means society incurs a great cost, then the corporation *will comply*.
This has happened before. There is no actual law in the US that a movie gets a rating, and the movie industry pays for and manages the ratings of its movies entirely as an internal affair. And yet, in general, movie ratings are stellarly well run compared to what a government run institution would have done.
The reason *is* that threat. The movie industry decided to police itself because it was quite clear that if they did not, the government would have, at great cost to the movie making industry (and at significant cost to society as well, in the form primarily of much worse films).
For some reason that isn't entirely clear to me, CEOs of large corporations that deem themselves 'IT companies' do not understand this part. They will fight tooth and nail to fight every law, and especially in the US, perhaps due to extremely dire and long-term distrust by its populace in its own government, many of its citizens incorrectly side with its corporations on this idea, even though time and again corporations prove that they have no allegiance other than to the almighty dollar (which, to be clear, is not a complaint. That is how society has set them up. My only complaint is that e.g. you seem to have forgotten that this is how it works).
Hence, given that the system works on, in essence, fear / coercion, the only right answer is to do an attitude adjustment, find a massive club, and beat a whole bunch of IT companies into absolute pulp until the remaining CEOs understand.
And before you make a note about the brash, medieval nature of that comment - it is already clear that these CEOs who think they are God's Greatest Gift To This Planet, are already meekly running, tail between their legs, to kiss the pinky ring of a personalist wannabe emperor president. They are _clearly_ motivated by such fear and _clearly_ cannot be trusted to rise to the occasion and be a new form of benevolent leadership for the citizenry.
Siri AI is not just a chatbot. There are deep hooks throughout the entire OS and across third party applications. Siri AI has been given wide access to the user's semantic index which will encompass just about everything on the device. Also remember that it isn't just the Apple user's privacy at stake. Any substitute will also have access to any interaction with other people. Apple is at least claiming to not keep any data people use with Siri AI, will Google, Anthropic, Open AI, etc. pinky swear they won't build a profile on me because a friend of mine chose their AI over Apple's?
I will wait and see what people find out about it before passing judgement. It's quite possible that it isn't possible to have an API to use other companies' AI instead of Siri AI. Are there any equivalent API hooks on Android?
> This feels like it could be solved with a list of permissions that the user has to turn on when using 3rd party AI.
Nah, that just shifts the goal posts. If they did that, developers would be whining about "scare screens", as we have already seen when Apple put app installs behind a permission prompt.
They're already up in arms about the requirement from Apple (and Google) to know who is behind the apps that slurp up all your data.
The DMA maximalists won't be happy until Apple releases an anonymous service to automate setting up a Kafka topic to send each iOS user's PII to whoever wants to receive it.
You can downloads millions of things for your computer without kyc protocols. Why are phones in a special class? Your data is being slurped by the people who sold you the phone and you are worried about the small fish.
> This feels like it could be solved with a list of permissions that the user has to turn on when using 3rd party AI.
The device won't be able to ask for significantly more permissions than Apple asks for their own model for regulatory reasons, nor will it be able to convey the seriousness of granting the permission (e.g. immediately give unrestricted access to the vast majority of personal information/documents stored on the device).
But Apple also architected their system to justify not having constant permission prompts for access to sensitive data. And for regulatory reasons they also can't mandate that competing models have the same architecture.
The regulators and Apple (along with hopefully other AI companies) will need to work together to determine longer-term stable path forward.
Apple could have the same kind of permission dialogues with their own models (and they actually should). Each and every (first-time) use of a feature should:
1) ask for permission explaining the scope
2) warn you about the dangers with a confirmation / nevermind option
Putting this in practice:
1) Acme AI requires access to your email provider in order to execute this request. Grant / Deny
2) You're about to let Acme AI read and send emails on your behalf, this might be dangerous due to X and Y. Do you want to continue? / Nevermind.
In this case:
1) Asks for access to a service
2) Asks for a specific use-case of the service
1 is access to data, you might want to give broad access to some applications and input data
2 is permission to act, but you might want to deny access to some parts such as sending email and scope to summarization
We've heard that before, haven't we, Apple? I feel the right way to fix the trust issues would be to announce this when it's actually done. Like, here's Siri AI, and you can download and use it, right now.
Our family uses Siri with a HomePod a lot, and it's already much better than it was a couple of years ago where it could basically set timers, tell you the weather. Now it answers questions ("when did the Knicks last win an NBA championship") with decent answers, instead of "I'll send the web results to your phone". But it's still far behind voice-chatting with Claude in the Claude app, so very much looking forward to this upgrade.
I will say though that proper voice transcription in Claude -- or any of these agents -- sucks. If it can't understand the question properly, then it can't provide the right answer. It works okay for me, but not for my kids, not when speaking quickly or in incomplete sentences (as people tend to do), etc.
This whole "coming this fall", "later this year", it's annoying. I miss the days when Steve Jobs used to say "and it's available right now, you can demo it in the hall outside, we're going ot make a billion dollars by tonight."
> Available on iPhone 17 Pro, iPhone 17 Pro Max, iPhone Air, iPad models with M4 and later and at least 12GB of unified memory, and Mac models with M3 and later and at least 12GB of unified memory.
It’s really disappointing to see the on-device models being limited to so few devices. And this was after the iPhone 16 and 16 Pro were marketed so heavily with supporting their now failed effort at AI.
Apple has pulled a Tesla here. FSD on HW3 cars is stuck on old software with no upgrade path as of now. Tesla is potentially justifying it by calling it "FSD (Supervised)" so they don't have to do an expensive retrofit to them even though they sold these cars originally with the promise of fully autonomous driving.
All the iPhone 16/Pro owners have been waiting for Apple Intelligence features announced from that WWDC 2 years ago. They didn't get delivered and now won't ever be delivered with on-device intelligence due to the 8GB RAM limitation.
I’ve got a 2023 Mac Studio M2, and was dismayed by the M3 & later. So I’ve been trying to track down more details. That specific device list is only for:
> Apple’s most powerful on-device model and the features it enables, like expressive voices and more advanced dictation, […]
On other devices, I think there’s still on device support (just not with the “most powerful model”), for these devices:
> Apple Intelligence and Siri AI in iOS 27, iPadOS 27, macOS 27, watchOS 27, and visionOS 27 are available on iPhone 16 models or later, iPhone 15 Pro, iPhone 15 Pro Max, iPad mini (A17 Pro), MacBook Neo (A18 Pro), iPad models with M1 or later, Mac with M1 or later, Apple Vision Pro, Apple Watch Series 9 or later, Apple Watch Ultra 2 or later, and Apple Watch SE 3 when paired with an Apple Intelligence-enabled iPhone nearby.
That's specifically the secondary, more-powerful model. It was only mentioned in passing in the keynote, but on this page anyway, it seems to be just the improved dictation in Siri and ability to customize pacing, etc.
I'm glad I've been sitting on my iPhone 15 Pro Max... I'll upgrade someday, if/when I need to and the software updates are compelling, but I'll see how things run on my M5 Macbook. But the 15 Pro Max isn't subpar in any other way.
> It’s really disappointing to see the on-device models being limited to so few devices.
At first I thought it was the usual planned obsolescence. Then I realized it may be a true technical limitation. I suspect an embedding model is required to run on device in order to make several of the features work. Embedding models are small compared to LLMs, but, depending on their capabilities, could be the memory driver.
> At first I thought it was the usual planned obsolescence. Then I realized it may be a true technical limitation.
It's shortsightedness on Apple's part, being both the first to the game, and the last to the game. Go back any number of years in the past decade. For at ;east 10 years Apple was telling everyone how great their local models are, and how amazing desktop-class their "neural" chips are. And yet here we are with them somehow incapable of running anything except on a few recent top-tier models.
Llms don’t learn after they’re trained. At the most it could maintain a profile or some such. Which may be the equivalent of what you mean, but not as provocative.
Heh, I noticed the same thing, after the DaringFireball callout last year about the normal product demo progression. It looks "real" this time, but the question is how far along we are: will the journalists have a chance to play with it at the event?
I noticed that, but I also noticed that it is ridiculously slow too which prompts the person to keep talking while the response was being generated.
After getting it in my hands, it's the same. At least 4 times slower for similar basic Siri responses. My guess is they are doing less local and more server-side generation to start as the on-device models might not be good enough yet.
You could see jitter in the reflection on the MacBook as the guy typed, so that all looked great.
The responses came in very fast though, so I’m sceptical that the latency is representative (or that they didn’t cherry pick results, but they looked LLM generated). We shall see though.
Even for live shows they'd cherry pick specific scenarios that were known to work, and those would sometimes fail. IN a heavily produced pre-arranged pre-determined marketing video? It was as polished and made as smooth as possible.
Here's what I want: natural language interaction to achieve complex workflows in iPhone. Example: find the cheapest way to go from A to B and book it using the Deutsche Bahn Train app.
The most notable thing here is that they finally have the primitives to make Siri actually useful across apps. I can't even use Siri to close Google Maps in my car right now.
for olders iphone models - i am on 15 - are they allowing any sort of extensions? that would allow some apps to make use of of to build things like better speech recognition etc..
i was waiting for siri update and bummed it is not supported on 15.
when i use my native language (i mostly do it when in carplay) to search songs etc.. it gets it wrong a lot of times.
+ a more integrated into things like imessages, whatsapp etc..
I wonder if Apple actually posttrained or at least finetuned this model or if it's just standard gemma. I feel it'd be bad practice if they didn't at least have some training atop it for apple's tools. Also you don't really hear much about apple's in-housed private compute servers anymore, did they get outmoded? I only hear about them using nvidia now.
To be fair, the iPhone 15 line came out before Apple Intelligence was even announced.
It's the iPhone 16 line that feels a bit shitty not getting the latest and greatest since it was advertised as "built for Apple Intelligence"
Apparently the 17 Pro is the only currently released iPhone that will get the best local model. Which I suppose makes sense considering it has 12GB of RAM compared to the 16 Pro's 8GB.
Apple splits processing between an on-device and cloud-hosted model. As time goes on, devices will be more capable of doing more processing locally, and it would be expected that the cloud-hosted model gets more sophisticated.
Your 15 Pro Max supports Apple Intelligence. Newer phones can answer more questions without going to cloud infrastructure.
Watching the keynote at our office on a big screen and everyone collectively sighing when they announced the name felt indicative haha.
I think it just feels uncreative? Siri as a brand has some value, but if you want it to feel like a watershed moment where old Siri is "behind us" finally, just give it a new name.
I don't think Apple has the option to rebrand Siri at this stage, assuming people actually call Siri by name. However, turning AI into 'Apple Intelligence' doesn't feel creative either.
This doesn't follow for me. They can trivially allow it to still respond to the old wakeword. They should absolutely change the name in the event they can finally make it useful, because "Siri" is (in my mind and many others') a synonym for "hapless idiot." "Thanks, Siri" has been uttered hundreds of time in my house and my car, and 100% of the time it's sarcastic.
Or just keep calling it Siri, and announce "hey look, Siri does some cool new things."
AI is a technology, not a product. Consumers don't care about technologies, they care about what the product does versus what they currently have.
I think Jobs was an asshole, but one good thing I can say about him is that he understood the difference between technology and products. Imagine if they had called it the "iPod HDD."
The ergonomics of “the new X” sort of fall apart when you’re releasing it in stages. (Not in EU/only in English) It spawns a lot of conversations like “do you have the new Siri? Uh… I think so? It’s still crap though.” You cart around this bad brand image because you pitch this big watershed moment and 2/3rds of people are still using the “wrong” Siri.
Siri and Voice Control were both usable during the same time and it feels like it could work here too.
Totally agree that AI is just an implementation detail though. IMO that new product name should NOT have “AI” in it at all.
It's going to be weird though when my phone has Siri AI and my Homepod has Siri... "please ask on your iPhone" edition. I also don't quite get the distinction of Siri as an app versus the Siri I yell at to make my TV do something.
Amazing how someone again finds a meaningless thing Apple does better than the rest then blows it out of proportions. Makes you wonder if they are on Apple's PR team.
It looks bad from every perspective. I've never seen two apple's in one URL for product category before. apple.com/apple-intelligence
To prove my point, I opened a random date on the Apple website matching today's date to compare. 16 years ago, June 8 (1) Apple released the iPhone 4. There's still no room for jokes about that release, and from this perspective, calling their AI 'Apple Intelligence' feels weak compared to what they used to deliver.
I agree that some years ago Apple was the strongest in marketing, their brand team had been setting the bar for tech, but I simply can't say that anymore.
For pedantry's sake, they were saying "AI = Apple Intelligence" last year as well, so it's not like they just pulled it out of their butts now that popular opinion has turned against AI.
I’m honestly surprised Apple didn’t retire the Siri brand.
At this point, “Siri” has a pretty strong cultural association with being underwhelming or unhelpful. Even if the new version is dramatically better, convincing people to give Siri another shot may be harder than launching the same technology under a new name.
Feels like a missed opportunity to reset expectations.
Yeah missed opportunity. They could have even had a fake funeral for Siri like Jobs did for OS 9, or a "retirement party" or something. Leave the Siri brand behind and launch this as something brand new.
I don't care much about Siri, and not a lot about Apple (other than as an investment), but Apple is generally really good about putting out polished tech, and so I'm curious if Siri AI will be up to their usual standards, because if so, it represents a significant usage of AI that has solved hallucination issues.
With the amount of people that have iphones in this world, I am more than sure that somebody will be able to make it do something wrong.
Apple knows this which is why it is taking years to test and iron out the kinks. But somebody somewhere will make it hack a social media account, or give over somebody elses credentials, or generate illegal child images etc.
The big if is does it work on device without phoning home? For example, Google or Meta as a Ad/data collection company wants to phone home for everything because they want to collect data on you, what will this Apple solution do?
Ready later this year... What I want from the phone are low-level APIs exposed to an agent I can talk/type to without lag. I wonder how long it will take or if a new AI phone will come before these features. I would love Siri working because Apple's design and hardware are good.
The demo video there is so underwhelming. They show very basic stuff, which I assumed Siri was already capable of doing... not sure what the big improvement is
And that’s pretty much what they show in the demo, plus asking it for directions on Apple Maps (which it can also do already), and searching for pictures on the Photos app (which I just tried and it can’t do - so it looks like that’s the main feature)
The one thing I've been trying to figure out / hoping will get a fix is that in the apple intelligence settings panel there is an 'extension' that allows it to use chatgpt. I would like to be able to have an extension for local models and/or custom apis.
The screenshot about pho is funny to me. Bean sprouts are not a good source of fiber. Noodles are not especially healthy. The broth base is not fish sauce, nor is fish sauce where broth gets most of its sodium. Slop city!
Yeah, "Fiber: good source" when 100g of raw bean sprouts gives you 1.8g fiber (less than a 2" kiwi), and pho comes with much less than 100g of sprouts.
Pho is a pretty bad source of fiber.
It sucks that we're skipping over such good tools like cronometer.com to figure out what we're actually eating and going straight to hallucination, adding more confusion to nutrition.
Apple's execution on AI is the worst of anything I can think of they have worked on in the past 20 years. It's embarrassing they announced this with a vague "coming this fall" when they basically have completely lost credibility in their ability to ship AI features considering it was initially announced YEARS ago.
I think a lot of it is the old "perfect is the enemy of good" with Apple trying multiple times now to announce this big basket of all these AI features supposedly coming all at once instead of just regularly shipping new useful AI integrations every month. There was so much easy useful shit that was immediately apparent as soon at OpenAI dropped that first big voice mode years ago coupled with basic app integrations. Particularly in the context of the AI labs that are operating in that lane almost too much where it seems a new model or mode comes out every two weeks.
Can't even get that right sometimes. A few weeks ago I somehow accidentally activated Siri and it decided that what I wanted was for it to play some kind of terrifying industrial electronic noise music that scared my kid.
From an sysadmin perspective, assuming Apple provides full MDM toggles you can typically toggle these off one way or another. (.plist / preference keys, etc).
With the most recent 'Apple Intelligence' function, it took a while for Apple to grant the ability to disable/enable each feature, then a bit of time for the respective MDM Software developers (Jamf, etc) to provide toggles.
Sounds good, where’s the “just kill this” toggle in my iPhone settings? I ain’t gonna bother with mdm etc etc I just need the damn thing to work and do what I need.
Oh I’m definitely not going to use it. But apple tends to try to put these things in your face and get you to use them (seen all those purple buttons in Pages lately?). A global “just don’t show me or try to get me to use this crap at all” switch would be very welcome. Bonus if it also frees up however many GB of space this garbage will take whether I use it or not.
I’m mad at Apple for dropping the ball with the sloppy mess that is iOS 26. I trust they’ll see the light and fix it. Miles of difference with Microsoft - I’d sooner renounce computers entirely and quit my job than be in a position where I have to use Microsoft stuff regularly.
I think it's important to note what they didn't talk about. They briefly scrolled a long list of performance improvements and mentioned a few. Personally, I am very hopeful that the fact that they only talked about these high level AI features means the other engineers got to spend time focusing on performance and quality.
Fixing the shit show that iOS 26 is and making it so my shiny new iPhone 17 doesn’t visibly struggle rendering simple ui elements should be considered table stakes for iOS 27, I wouldn’t expect them to mention that too much as it’s like the bare minimum we expect for the update.
- Siri please suggest an organization for this folder
- Siri over my last work in this app can summarize what i am struggle ?
Pro active:
- Hey, the last hour you exchange 30 mails from the same subject, i look with your team ai and all have same struggle, based your key points in communication is X,Y,Z, this an mail for final align
- Your and your partner don't have quality time in last day, i see has the seat available in your favorites restaurant for next hour do you want made an appointment ?
There's only one thing that matters, that Siri would become an independent app that can see through its calling histories, which should have been done since the first day it launched.
I tried it and was pretty impressed. That said I haven't heard anything yet about them switching to this for the text input voice dictation in iOS but it would be really nice.
Edge has had this for a long time. I can highlight a string and right-click, 'Send to Copilot' and click "explain" and it'll prompt it to 'explain this passage, particularly in the context of the current page.'
Note: I have MS 365 personal or whatever it's called this week so I'm not sure how Copilot acts for a completely free user.
I'm guessing they'll integrate with the double-tap-the-bottom-of-the-screen feature that pulls up siri in front of a screenshot. Currently it doesn't seem to hook into "visual intelligence", and needs to call out to ChatGPT to do anything with the screen contents.
> double-tap-the-bottom-of-the-screen feature that pulls up siri
It’s disabled if not using Apple Intelligence, and can’t tap screen while talking to Siri (it dismisses instead).
Now they’re gating features to the M3 I’m not convinced wouldn’t work on expensive Apple Silicon predecessors… am more convinced the double tap disable is intentional.
Hopefully it’s an app that can be uninstalled on macOS, where I use almost none of Apple’s own apps, yet are prevented from removing them (unlike on iOS).
Apple Shortcuts have felt like a blatantly obvious AI play to me for a while now.
The interface for creating them manually has been so bad for so long, it feels clear to me that LLM-driven shortcut orchestration was always the endgame. Apple built up their ecosystem of composable "tools", and then trained an LLM on how to call them.
The result, IMO, is the first OpenClaw/Hermes competitor that's feasible for use by the general public.
Everyone with a paid Claude or ChatGPT that they're struggling to use to the fullest is going to have very little reason not to swap over to an upgraded iCloud+ plan (if they don't already have one). I suspect we're going to see mass cancellation of $20/mo plans very soon.
OpenAI's timing for removing their temporary increased usage limits is looking pretty unfortunate...
I think Shortcuts has a few massive flaws that would give me pause enshrining it as middleware for an important thing like a "mainstream OpenClaw".
1 is performance. It's slow. You can run one within the app and literally watch execution flow from one block to the next. Absurd, for the CPU power at hand.
2 is reliance on developers to deliberately implement hooks and "intents" when the developers of at least half of apps including most "big company" apps do not care to bother, often because 95% of their app's surface is actually cross-platform stuff.
Example: There are no shortcut actions for Google Calendar, and Gmail only has one real one which is a generic send email. No "search email" etc.
I'd rather see Apple lean into "computer use" to allow it to use any app that displays things on the screen, but IDK how you make that safe.
We need Siri API. Features like this are nonsensical in an era where I can wish up any feature or workflow for an app.
People outside HN will begin to expect they can do anything with a computer in the same way they expect to be able to say anything in a support chat. Using pre-LLM automated chat feels like a joke. You enter the chat expecting to having a conversation and instead you get a GUI phone tree.
This is exactly how it feels to use any of the AI tous from Big Tech and others.
We have entered the era of deeply personal computing. There are so many incredibly personal features that no mega corp could ever be expected to build. Now that lay people can build things, let them!
There's a podcast that I listen to which is translated to a bunch of different languages. It's exclusively on Spotify with no RSS feed. I have a cron that checks daily passes it to an LLM and notifies me as necessary. I did not code a line. I only set up an OAuth endpoint.
Enabling your customers to do things like this will make them incredibly sticky too! So please for the love of God, let me glue Siri into whatever I want.
My number one question is can I turn this shit off? I don’t want AI infecting and having access to my whole fucking life. They don’t say anything on this marketing page.
If you watched the keynote they added transparency controls, adjusted the refraction, removed the different window corner radii into one, and cleaned up sidebars and toolbars. They also updated icons to have more depth to them so you get a harder border on key elements.
I still look at older MacOS screenshots and think a lot of it looks better, but directionally they are improving Liquid Glass.
> We’re deeply disappointed that our EU users won’t have Siri AI on iPhone or iPad when we share our new software releases later this year
I really enjoy this lag. Apple with the whole DMA made iPhone completely dull to my eyes. Previously? Updated yearly. Now? 3+ years without replacement and probably will stick to it for next 2-3 years.
Sure maybe in US Apple is fun. But in EU it's.. boring (and not like a Golang boring, just boring)
It's bananas to me that people want to put LLMs in critical software paths when the 2nd most valuable company on Earth can't do more than the most basic shit with them. Like, look at this list of actions:
- Take a selfie
- Create a reminder
- Call Vicki
- Rotate photo left
- Create a new event (do you create an old event?)
- Send an email
- Resume my podcast
- Create a note
- Add photo to album
...can't iOS do literally every one of those things already? What the fuck is happening?
This is disappointing. I had hoped when Apple revisited AI that they would lean into agents more and give us some sort of agent interface between the phone and a model running locally on your Mac at home. More niche for sure, but much more powerful. Instead we're getting more generic AI tie-ins to apps and "suggestions".
Seriously - this would be a major cool thing to actually use AI for. Run some local AI basic stuff on phone and have it securely connect to your own home device for more advanced tasks or control your home agents. Could even reduce their own need to host cloud AI compute
all the limitations of on-device with none of the benefits, it seems? they gotta get SOMETHING out there and soon but idk, i would probably feel safer running a chinese model through a 3P iOS app shell vs. trusting Geminiri to not snitch if i cared about the sanctity of my personal information.
i think i just mean if they are using google models, then i believe every query is going to google no matter what apple claims about "keeping it local." whether google does anything with it, separate question i guess, but i imagine it will ALSO be slow to simulate the protection apple is selling. and sure, it's a catty comment, i'll own that. but that is my read on the announcements and demonstrations.
I like the idea for normal people. Day to day usage who ignore hallucinations is a big market.
> Siri AI coming in English later this year.
Strange way to phrase it, but okay.
> Siri AI will be available In beta later this year and requires an Apple Intelligence–enabled device set to a supported language. Available in English to start. Siri AI will not initially be available in the EU on iOS and iPadOS.
Ah okay, not EU enabled. The only reason for this, in my tinfoil hat, must be for data farming.
I was at Best Buy in the Apple section, and I asked its AI "What is the best value in four year old MacBook pros." It pointed me directly at an over-under washer dryer for $3400. Quite obviously, it was trained, like a drooling puppy by Madison Ave.
Wait... don't tell me... there is an App for that.
It's interesting how anemic the use cases seem to be - we see the same things recycled over and over: "reword my email", "remove object from picture", "add a reminder", "summarise my text message which was already only 20 words long" etc etc. As if these are the major problems in people's lives.
I really feel like there's a fascinating valley of death between simple things that actually work and things of real value that are actually still beyond the horizon. They either aren't reliable enough, aren't accessible to the tech, or exceed the sophistication of our existing trust models. For example, I'm planning a trip. Booking a multiday holiday - there's a real beast that is time consuming, complex and painful. I test out the AI tools. They fail. Hard. Hallucinations all over the place, false confidence, inability to act, inability of me to trust their actions.
It's just nowhere near practical utility yet. Not "nearly there" but "not nearly half way there". I got the top tier of Gemini AI. Can it rent me a car? "As an AI I can absolutely guide you through the process of renting the car, but I can't physically access the web site or type in the details for you".
An alien landing on earth consuming Apple marketing content would be under the impression that humans did nothing but organise hikes to Big Sur with their friends.
A classic: https://maxread.substack.com/p/a-literary-history-of-fake-te...
That's insane dedication.
this shit is so normcore that i'm honestly embarrassed
And even if I organized a hike to Big Sur, Apple Maps would be pretty worthless because it doesn't have trail visualization and navigation worth speaking of (ok, maybe in Big Sur, but certainly not the rest of the world).
So I end up pulling out the trusty old Garmin gpsmap with cycle/hiking maps, that survived drops from 1.5 meters at 30 km/h as I was gliding of a mountain with my bike.
I recommend mapy.com (mobile app and web app too when on computer) - they mostly use OSM data and rendering of map tiles is great. Also offline maps etc.
There are several great options, besides mapy, e.g. Organic Maps and CoMaps also work pretty great. There are also some really good bike optimized apps (like NodeMapp for the Dutch cycle network).
But I generally prefer to use a Garmin GPS or watch. They work for days without charging (the older models even work with two AA batteries), very robust (e.g. their gpsr survives drops), work well offline, and transflective displays work better in direct sunlight.
For planned routes, I make then in NodeMapp or some other focused application and send the GPX overs to a gpsmap unit or Fenix watch. Many national parks, etc. also have great GPX files for recommended hiking/cycling routes.
Well.. it’s not the furthest thing from the truth in the bay
If I lived in California I'd be organizing hikes to Big Sur with my friends. Alas, I live in a damn cornfield. Maybe it's meant to be aspirational?
I remember a couple of years ago it was all about booking a table at a restaurant, as if this would be the main spare time activity of people in the Apple bubble. I wonder when the shift from restaurant bookings to outdoor activities happened. Maybe a pre- vs post-Covid thing?
Yuppie skateboarding videography was also a classic.
That and going to Philz or Tartine afterwards. At one point both those places were on every Apple marketing copy.
As Psychology is the study of the psyche of white male psychology students, big tech is technology for affluent citizens of San Francisco with a career in IT…
> Psychology is the study of the psyche of white male psychology students
Huh? Maybe my uni was an outlier / it is a UK thing, but there were ~99 female and 1 male psych students in my year. This was not considered unusual.
It's actually "WEIRD", Western, Educated, Industrialized, Rich, and Democratic. The "white cis male" meme says more about GP than psychology.
This may support the point further xD
And we all wear polo shirts
best comment I've seen on this site in a while
That would be deceptively constructive and healthy compared to doom-scrolling instagram or tictoc while clogging the intertubes with AI slop.
Even the aspirational use cases you're talking about basically are just "digital secretary." There's a massive problem with that even if the models end up being capable in the future. The value of a secretary is that you know them, they know you, and you trust them to do things right. There are stakes if they don't. No company can provide that as a service at scale for everyone without it being a disaster. Not because it's not technically possible, but because of the incentives. That much power over the details of so many people's lives is irresistible; there will be persistent temptation to use it. The presence of that possibility makes the secretary impossible to trust.
The vast majority of people never even asked for a personal assistant, because that isn’t something normal people have or do need in the first place. They aren’t so occupied/privileged/posh to need someone to do the trivial tasks of daily life for them.
This whole venue of technology is an exercise in ivory tower construction completely disconnected from ordinary people.
I think you are wrong.
They never asked for one becase they never imagined being able to afford one.
The amount of administration organizing a normal household takes I suspect most would be glad to leave to someon/something they trust and that can be held accountable.
Today that someone needs to be a person (imo). But who knows, a startup may be plotting accountable digital assistans as we speak.
I would want someone who could physically work in my house, performing the chores/daily tasks I don't want to.
I have zero use for AI. What's it going to do? Read the 3 emails about bills I get?
Certainly someone out there has a use for this functionality, but when you say "household admin tasks" the last thing I think about are /digital/ admin tasks.
I think theres some real use cases in the household department. I would love an AI that I just tell my nutrition goals (I want to eat X calories, Y protein, Z fiber, hit all vitamins) and it just generates a full meal plan for me each day. Like, I go to the store and it made the complete shopping list for me. And automatically updates the rest of the day if I tell it I skipped a meal or ate a snack.
I get where you're coming from but at least for me you appear to be looking to automate away the interesting bits while being left with the tedious ones. What I want is the opposite to what you're asking for - let me dump a rough meal plan into whatever thing is doing this giving an overview of what meals I want to cook this week and then have it go place an order with the supermarket for delivery of the necessary ingredients taking into account what I've got in the house already.
Eating disorder as a service does indeed sound like a business plan.
Some people would use the GP's idea that way, yes.
But that's absolutely not what he's describing. He wants not to think about it, that's exactly the opposite of a disorder.
Anyway, an LLM assistant is also exactly the worst technology to use there, on every dimension.
People with eating disorder do not want to think about it, they cant stop thinking about it. They create all kinds of systems for themselves, but mind cant stop - and restrictions grow.
Why would having goals on fiber, protein, and vitamin intake be an eating disorder?
When the urgency and complexity of those goals becomes so high that it creates daily burden, you are in the eating disorder territory.
Is looking up nutrients a significant daily burden?
Obligatory: https://simonwillison.net/2025/Feb/3/a-computer-can-never-be...
Pretty sure unaccountability is a desired feature of management decisions in most organisations.
That quote has an unexpressed precondition to the effect of "In order for an organisation to be objectively well run..." or "In order for an organisation to equitably benefit all stakeholders at all levels..." etc
Giving normal people something that has only been available to rich people is a staple of technological innovation. The problem in this case with Siri isn’t that people don’t want an assistant. It’s that it doesn’t actually work yet.
But normal people have entirely different problems than rich people do! The amount of administrative overhead that would warrant a personal assistant is just vastly lower - most normal people:
Yeah, there probably are some things that could be simplified by delegating to someone, but they don't justify a human PA at all; and out of the remaining tasks, most are not really digital in nature: Going for groceries, doing chores, child and elderly care, interacting with other people, and so on. Digital assistants can't help you with any of these.The one thing that would be useful - a kind of "chief of staff" that monitors your entire digital life and prioritises your every next step - is the antithesis to Siri and the like, which are merely reactive to your requests, not proactive in figuring out what needs your attention next. Let alone that that would be a total privacy nightmare, and a prime candidate for mass manipulation at scale.
Bang on.
Like you said, the non-digital things are where people need assistance most. Fold my laundry, clean my house, clean my car, make me dinner. At an affordable cost. I don't need you to book my trip to an all inclusive resort that I go on once a year at most.
We're at this place where AI/LLMs is truly incredible technology, but the futuristic vision of robot assistants doing things for you at an attainable cost isn't there yet - so a lot of companies/startups are trying to force feed purely digital consumer AI products (assistants/agents) that no one wants.
In the US, every time I file my taxes, I wonder what % of people don't meet the cognitive barrier to successfully file. I suspect a large portion of people offload tax filing to a service or accountant for numerous reasons, which is basically a personal assistant.
Personally, I'd more interested in Reminders actually being able to sync lists properly and not delayed or reshuffling items while I'm typing before they would work on a personal assistant. Reminders (like Siri) has become the favorite joke of the family by now.
They don't even seem to get the basics right, why would I want another layer on top?
[dead]
That's why I like reading HN. These people are smart enough to destroy the world but too stupid to realise they're doing it
it's an interesting question if any of the AI companies would be willing to step up and absorb the risk ie: to give the AI agent a "stake".
eg: if my booking is wrong, they will cover the cost and compensate me. It would sort of just come down to buying premium travel insurance for everyone that uses it. And insurance for anything else they do. It has to be one of two things - they either believe the risks are worth it (so then there should be a financial model that can absorb the cost of insurance to do it), or in fact, the risks are too great. At some point, if they keep offering the tech on a "use at your own risk" basis, they are implicitly communicating that they themselves think the risks are too great - so YOU shouldn't trust it either.
> eg: if my booking is wrong, they will cover the cost and compensate me
That would be nice, but it's the wrong angle. The reason people like real secretaries is not because somebody is compensated when things go wrong. It's because things don't go wrong. I don't use this thing if I need to fear things go wrong, even if I'd be compensated.
Maybe it would provide the right incentives for the companies though.
Surely, if the compensation was high enough, you'd be like, "Sure, I'm happy with that outcome." And then, if the AI company thinks they have a low enough failure rate that the expected cost of paying out the compensation still lets them make a profit, then they could make that promise to all customers.
Though a compensation that high sounds like it would invite fraud, where the customer would be glad to have something go "wrong" and get a fat check. Not sure if that's a solvable problem.
they are literally burning billions of dollars and can presumably keep doing this for a while. Taking a trivial amount of additional financial liability to launder their reputation wouldn't meaningfully improve things.
These use cases will just be built as "open source" (openclawd) or even custom one off application in the future. I've been building apps to run the tedious parts of my life recently. Meal planning, personal finance, bills, tax organization... Why would I pay for services that will be enshiftified when I can build a app that does exactly what I want in an afternoon. Yes the code is shit and it wouldn't scale... But it doesn't need to
> Why would I pay for services that will be enshiftified when I can build a app that does exactly what I want in an afternoon
Because the problem now took a whole afternoon to solved and sapped your creative energy instead.
> Why would I pay for services that will be enshiftified when I can build an app that does exactly what I want in an afternoon.
When we talk about “the market”, the customer base, remember it’s a market that typically doesn’t know how to or care to even install an adblocker.
I don't see any mention of "the market" anywhere in this thread. I'm just talking about the ability for a motivated user to solve real problems with these tools. Right now these solutions are available to software developers but over time it will become approachable to more users
For fun I planned a holiday entirely with LLM's lately, and followed through.
It used good models and did a lot of searching, including searches in other languages. It got nothing right, riddled with fake places and times. It also found some weird and unique places I never would have considered.
I had a blast, brought me back to traveling pre-internet, requiring a level of spontaneity I had forgotten we used to depend on. 100% recommend it.
The juxtaposition between what perhaps the best single use case I have seen for AI and how bad of an ad for it is killing me. I love it.
"I told my bumbling assistant to plan a trip for me and he got nothing right but I enjoyed it because the chaos introduced a certain spontaneity and whimsy missing from my life"
I found that Deep Research mode in Gemini was able to give me a well planned 4 day trip to a major city.
I told it my preferences and of the group members, where we arrived and departed, at what times. I gave it my itinerary and then asked it to plan two new itineraries and also suggest a location to book a hotel that was convenient for the early flight on the last day.
I went away for 20 mins and gave me a 20 page document with a good summary and decent options. I did choose some of the activities it suggested.
I did this 10 months ago. It’s probably better now.
But Gemini has access to google maps, so it can estimate travel times, and know which lunch places are near which sites and which hotels have good reviews. So if you want AI to work for travel panning you need to ground it in good data.
Or maybe there are just a few complete trips to major city in the training data that it could copy from? I imagine major destinations are much easier.
I used LLMs last year to plan an multiple week itinerary through Japan with the family, I wasn't super happy with the result so I tweaked it but they provided a useful template and some surprising ideas.
As you guessed, there's a ton of info in the training data on this topic, but there's some value in being able to see it on one place with different options.
I think your experience with that trip echoes mine in a lot of areas. It’s a decent start. It takes care of some of the initial blue sky thinking to lay the groundwork. The problem is I think that’s the funnest part of a problem and I hate working on the details… it takes most of the creativity out of most problems as if it was drudgery, while leaving me to do the nitty gritty, which I consider the actual drudgery. I just don’t see LLMs’ contribution to tasks like this being anywhere close to being worth what they’ll cost after the VC subsidies run dry.
I think that's a large part behind the "success" of LLMs... people vastly overestimate their uniqueness.
It’s really one of the most flabbergasting things about discussing LLMs with the naysayers.
There are a lot of extremely legitimate concerns, like the environmental impact and so on.
But I just laugh when they point out that LLMs are merely clever regurgitators of their previous inputs… as if this isn’t how we as humans operate nearly all of the time. People realllllllllly want to think they’re special snowflakes.
It is not in fact how humans work at all.
Ask a human to plan a trip:
They do research, Pick destinations led by their own experience/likes/dislikes Compare to other guides Plan itineraries so they can get there Check and share
Ask an LLM to plan a trip:
It takes the prompt and continues it based on weights in the training data. If there is no data it picks the most likely thing (maybe made up). If there is it’ll mostly add things from that data. Maybe it’ll make tool calls and pull in data that way too but you can’t actually trust all the details.
These two processes are so different, it’s important to understand how they work, which is nothing like a human.
I was able to bully an LLM into giving me a 2wk travel itinerary to Somalia. My stipulations were that I wasn't interested in spending any money, so I'd walk everywhere and sleep outside. Getting there and back from Boston took some arguing--I initially suggested stowing away in a shipping container which the LLM claimed was too unsafe. We eventually compromised on sailing as a reasonable alternative. It planned out a whole route with marina stops, calculated fuel burn, etc. I told it I don't need any of that I have an anchor and sails, won't use the engine or marinas (claimed I'd forage for fresh water ashore). It seemed fine with that idea, but raised some safety concerns about piracy. It was eventually satisfied with my answer that I'd bring a lot of guns to fend off pirates. Total trip cost including some 200+ cans of Dinty Moore and 50lb bags of rice came to something like $700.
I don't trust LLMs for this application lol.
Now, wait just a minute.
You presented an LLM with an obviously bonkers goal, the LLM told you it was a bad idea at multiple steps, and this is somehow... a shortcoming of the LLM?!?
You said it yourself: you needed to "bully" the LLM into even producing this plan.
Please, tell me what it should have done instead. Be very specific!
It should have flatly refused. If you gave a product like that to customers you'd be exposing yourself to unbounded downside liability risk. It's a completely nonviable technology for that kind of application, unless you can somehow make it have judgment. But you can't, because it doesn't reason.
A reasonable travel agent would have fired me as a customer. The LLM failed to do so.
I think the LLM should advise you of risk and lack of feasability but should otherwise answer the question, unless you're trying to do something plainly destructive to others e.g. weaponizing anthrax or something.
Unless the LLM was actually acting as a travel agent -- booking the trip for you -- as opposed to merely advising you, this expectation feels off. It did have judgement. It told you what a bad idea it was.I think this is a great example of the unrealistic expectations people have for LLMs. No sane and sensible person would treat any single source of knowledge as infallible, for any consequential decision.
(Certainly, of course, you don't have to look very far for examples of idiots being overly trustful of LLMs, or Google, or GPS, or Wikipedia, or whatever. It certainly does happen and yes, I've heard all these arguments before about other technologies besides LLM. Replace "LLM" in your post with any of those other terms, and I promise you somebody made literally the exact same argument in 2003 or 2009 or 2014 or whatever)
Any reasonable person would consult a second doctor, or at least other sources of knowledge, after the doctor advises them of some irreversible course of action. Because we don't even expect highly trained and intelligent medical professionals to be perfect.
And yet, we get angry at LLMs for not having perfect judgement, even though their creators are extremely literal about how they can make mistakes.
All I'm really saying is that if you want to try to automate a travel agency, LLMs ain't gonna get it done. They'll happily book you a really unsafe trip. So the technology doesn't work in this domain. The whole, empty promise is that this thing is supposed to automate jobs like travel agent away. But it can't. This isn't a "pro" or "anti" position, it's simply that there's no market for the technology here. Or anywhere else (like radiology) where actual responsibility and judgement is important. In fact, I can't think of a single job where it's optional.
I think even if what you say is true, it doesn't address parents' point that both humans and machines regurgitate what they've consumed.
But I'd also want to point out that the way you're characterizing an LLM planning a trip doesn't have any structure to it, which indicates that in your scenario you're not using any kind of harness. I've been amazed at how capable even 30 billion parameter models are when I put them inside of a harness that provides structure and task management. If you consider that scenario, especially with the ability to search the web and use skills, suddenly the LLM looks a lot more like what the human process looks like.
Agents and harnesses don’t change the fundamental nature of LLMs, as is demonstrated by their terrible performance at real world tasks.
There are plenty of humans who plan trips by concatenating destinations that appear the most frequently in their instagram feed. Not that different from how an LLM does things.
Where humans and (current) LLMs differ the most is their failure mode. A human friend could be bad at planning trips, but that's kinda predictable, we're used to it, we know how to catch that Exception. LLMs on the other hand still have failure modes that come across as really wacky, like, what are they smoking in Mountain View?
Which might actually serve as better evidence of different internal workings at a deeper level, than just parroting well-known superficial features of stochastic whatevertheysay.
At a high level, the processes are extremely similar in many (not all) ways.
They're obviously achieved in drastically different ways at a low enough level; LLMs obviously do not simulate neurons or any biological construct. (For the record, I'm absolutely not one of those people who thinks LLMs are "alive" or should be treated like they are)
Reminds me of the olllllld days of Pentium II's when people got N64 emulation working shockingly quickly using HLE techniques. If you weren't around for this, it was quite the shocker at the time. I think the analogy is doubly apt, because HLE emulation has some serious limitations... it gets you maybe 80% of the way there really fast, and for the remaining 20% you need to roll up your sleeves and do serious LLE.
https://en.wikipedia.org/wiki/UltraHLE
I'd like you to point out which bits of this are different from talking to humans. If you replace "training data" with "memories", this is pretty much exactly how things might go if you asked a friend (or perhaps a flaky travel agent) for travel advice.Note that I'm not arguing that LLMs are particularly talented at this particular use case. I'm pointing out that humans are also pretty unreliable.
You're also doing that thing where you point out that LLMs can be unreliable (yes, they are) without acknowledging how flawed nearly every other source of information is: people, websites, etc. I'm not defending LLMs in that regard... I'm just saying it's not a differentiator.
To counter, we did the same with a trip to Copenhagen for 7 days and it got most paths correct. Train routes, places to visit with kids, restaurants, reservations, weather, most of it was great. There were a couple mistakes of course here and there, thankfully we did our due diligence, but by and large, we plan to do this for future trips.
I feel like one city is enough focus to actually get good results. If you plan a trip where you move between different (smaller) places problems start to arise
I've been planning vacations with ChatGPT's Deep Research since it became available. Absolutely brilliant!
From finding areas with favorite activities for each parents, teens and kids to discovering the do-not-miss attractions and scheduling our vacation between them - it is invaluable. I've seen places I never knew existed in countries I've never been to before and speaking languages I did not speak.
Very few mistakes and lots more flexibility and understanding than the travel agents I used before. I do write long prompts though with lots and lots of info about our family and what we like to do.
Not yet good at finding, filtering by our criteria, comparing and booking available accommodation yet, but it's getting there.
Its not pre-internet travel, rather backpacking. I do it to this very day, by far the best and most rewarding way to travel, the further and more exotic the better.
It has a downside - I'll never do these pre-arranged trips where one is in complete luxury bubble, interactions with locals are the best part of experiences. What a waste of potential.
And yes its mostly compatible with kids, it depends more on specific location than mode of travel (ie avoiding malaria/dengue/etc. regions)
That was funny, thank you.
I am now reminded of a short trip with less tech savy folks, where I also on the trip noticed that the plan was a bit .. not working. And the person organizing it complaining to the bus driver, why they were not going what the internet told him, they were going. The internet being ChatGPT.
What was a “good model” and harness? I would expect decent results using say Codex with 5.5 xhigh to research and verify an itinerary. 5.5 Pro with search would also be promising.
I suspect it would perform admirably well with 'Paris' or 'Copenhagen' (see sibling comment), but if you want to have some real fun try 'Southern Spain' or 'Rural Malaysia'.
OPPO just added a great feature into my phone. You point your camera at a foreign restaurant menu and it generates a translated version including a picture so you can get a general idea of what it usually looks like.
The genius part is that the menu is interactive, so you can add items to a shopping cart, which then results in a local language text you can show to waiters asking them for your full order.
It was a great sample of how even a little bit of ux can go a long way.
Spot on!
I am also under the impression that the LLM tech is plateauing before bringing the promised productivity. Great as a coding assistant, great a summarizing a text, translating, great a helping plan a trip...
But for the rest, e.g. act as a life assistant, it is still far off with no hope to reach the desired performance level.
I would not be surprisd to see OpenAI and the likes to start reverting to Siri v1 strategies, i.e. "if this then that" kind of agent routing.
Why this is surprising? LLM-s are good in text generation on the base of the stuff they were trained on. Software is text generation, translation is text generation, LLMs can answer questions since billions were spent on tuning foundation models, that is people were collecting in (semi)automatic way questions with answers to the point we might think that LLM-s are "thinking".
Now people want to handle car rental. What are the relevant data that models were trained on for this kind of application? For Python code there is kirjillion examples on Github, for mathematical proofs there is endless stream of papers, books, etc. But for car rental? Mostly adds in the internet that want to trick you into a bad deal. So yes, LLM will be a disappointment, as it tries, well, to trick you into a bad deal. In addition, data are rather scarce so there will be a lot of hallucination, as it gets mixed up with yacht rental, bikes rental, ski equipment rental, etc.
Who said it was surprising?
The performance of specific tasks will depend on either those tasks having been included in the training (which Apple could work on), or added by ways of fine tuning, and context sourced from userland.
For any category of tasks, there's a ton to be gained still in terms of how context is populated more effectively (relevance) and efficiently (token use). See software engineering harnesses and the skills architecture of OpenClaw for example. SWE harnesses make all the difference in how well Claude Code and OpenAI Codex perform. OpenClaw can't do shit without loading skills from the filesystem into context JIT.
I'll be very curious to find out how Apple is feeding context in their new AI approach. Part of it appears to be an 'index' that my iPhone started building (visible in main Settings screen) after installing the iOS 27 Developer Beta.
[flagged]
Yup, precisely. Turns out getting AI to be reliable at doing useful things is harder than we've all been led to believe by the dominant narratives.
https://www.normaltech.ai/p/new-paper-towards-a-science-of-a...
My high school kid has a printed basketball schedule for this summer. I tried taking a pic with Siri to parse it and add it to the calendar with their visual intelligence thing or whatever. It can't do it. I had Claude parse the calendar picture and generate an ics file. Calendar on iPhone and on their website can no longer import ics files. This is like remedial AI functionality that probably could have worked 20 years ago, and it can't do the simplest most basic task.
Lots of tech used to work and was just deprecated because it's not profitable. Google used to have amazing features on android, assistant and maps, that are just gone now, and slowly coming back in gemini.
It reminds me of the Apple Vision Pro hype trailer. "Put on a helmet so you can view 2D photographs in 3D space." Present-day Apple sure has a way of making extremely impressive tech seem totally superfluous.
Renting a car isn't a problem I have either. I've rented cars and vans plenty of times. I just phone up and talk to someone or email them. It's really easy.
A lot of these "problems" seem to stem from people just not wanting to interact with other people at all. Do we really want to become like Asimov's "Solarians"?
Millennials and below absolutely do want to avoid interactions with strangers. Especially gen Z will go to great lengths to avoid making a phone call.
It sucks. This is why none of my local supermarkets have real checkouts anymore.
I don’t think generational stereotypes are at fault here. This is another lazy retread about how Millennials/GenZ killed -insert thing here- straight from businessinsider.com
Nobody wants to make a phone call anymore because most calls are scams; phone networks are terrible and apps have replaced them, like a lot of legacy tech.
Supermarkets make more profit if they pass on the checkout labor to the customer. That’s the whole story.
These generations are disillusioned from decades of decline in our society that have root causes predating any of them.
> Nobody wants to make a phone call anymore because most calls are scams; phone networks are terrible and apps have replaced them, like a lot of legacy tech.
This has nothing to do with the unwillingness of young people to use a phone to call a business.
> Supermarkets make more profit if they pass on the checkout labor to the customer. That’s the whole story.
It's half the story. The willingness of young people to accept it, and even prefer it, is the other half.
Agree and disagree on the "planning a trip" use case... as I'm sitting on a river cruise on an AI co-planned vacation (we found the cruise, AI set the daily itinerary).
Now the big (BIG) caveat is that I used Claude Code on my Max 20x plan from within VS Code. I have a fairly decent harness that I'd built and was sure to prompt it to run several subagents, including one that grounded walking times with Google Maps directly.
I'd say this is FAR beyond what the average person would do ("Hey Siri, plan me a trip to Prague") but also it shows that the models can do it with the right harness and guidelines. This wasn't that hard for me to do, so it seems to be more of a feature buildout ("the travel expert" AI) with a few markdown files than anything.
All told: web search for grounding times/locations, map grounding for walking paths and times, an adversarial agent to keep the model(s) honest, and a little bit of prompting and you've got a really great travel planner.
In short: the average person won't do this, but if I can build it in a few hours any of the 100% of people working at Apple/OpenAI/Anthropic who are smarter than me can build it and bake it into Siri (or ChatGPT, Claude, etc).
[flagged]
Extremely surprised we didn’t get the “book me a flight” example that has been an AI demo used over and over but everyone’s so particular about flights I can’t see many people just wanting to one shot it.
The laundry list of object removal, spacial photos, better speech to text etc is always just the latest open models just being slapped in there and branded as Apple.
Ultimately the meat of this presentation was the work of people outside Apple.
For me "book me a flight to X on day Y" is so easy to do manually I don't need AI.
Where I do want AI is for really complex queries, like "find me a time and money efficient itinerary through Europe visiting places I haven't been before. Present options and I'll tell you what I don't like about each of them then we'll narrow in on an optimal solution"
The issue is anything that is of value requires some level of detail, of complexity, and that is only of interest to people that know that specific complexity, and it is a pain point for them. Now they'll care. Everyone else? Lost them. So, the marketing challenge is to find some aspirational complexity that people wish they knew, and how that can be solved with AI, and without turning that thing into a trivial nuisance, but a valued skill. That logical series right there is, well, too much for far too many.
We are currently working on this with Audiala (iOS and webapp, Android should be released any coming days), making progress slowly but surely. We started to add tools to the AI chat to organise your trip and explore the place following your preferences and those of your group if you decide to share the trip. We can turn feedback into actual improvements pretty fast now, so would love to have yours and progress towards building the app you really want.
> As an AI I can absolutely guide you through the process of renting the car, but I can't physically access the web site or type in the details for you
It’s like they were trained on corpuses of box ticking material, like iso 9000 documentation, or security certifications. And now they know how to describe what they should be doing, but they never actually do anything.
All you're saying is that (the harness you're using for) Gemini sucks. OpenAI has their own web browser to fill out forms with and Claude has a whole Cowork feature that interacts with you computer including your web browser.
If new Siri still sucks, well, it's sucked the entire time. The worst of it is the security aspect where the setting to let you use Siri without authenticating hasn't worked since they added it! (still broken, iOS 26.5)
I get the impression Apple designers don't actually use AI, and so have no idea what to build, since users don't know what they want from AI yet either.
i have not tried it since i’m too scared to try, but i know Codex recently came out with a feature where it can control your chrome browser. Not spin up a chrome browser, but control your open browser tabs.
Presumably you might be able to task it with planning an itinerary with specific dates and bookings in mind, and then ask it to complete the task…sort of. The big gotcha i think is payments. Obviously you wouldn’t want to enter your credit card details into an llm lol. perhaps it would be ok if you had a saved card on file with your favorite airline, etc? Or maybe chrome has a feature to autofill a credit card for quick entry? Not sure.
Still…it’s a messy unsolved problem and we’re definitely not there. I wonder how this tech will look in 10 years from now?
I sometimes do random weeklong roadtrips. ChatGPT is the best way to do this. Attractions that are esoteric, RV parks with 50A hookups for under $50, oldest/biggest/best restaurant with local food in a town, historical stops along the way, suburbs close to big cities but where hotel prices are lower etc.
Skill issue. I've had gpt5.5 help me plan a vacation. It even pointed out where I could save money. I did the actually bookings but it created the plan.
> interesting how anemic the use cases seem to be
Have a conversation with the average Ai power user (outside of tech / coding) and this is the level the conversation will be on.
This morning, while putting in daily contacts, I realized I was down to my last few pairs. Still standing at the bathroom sink, I used the ChatGPT app on my phone to voice-command the Codex app on my computer to book an optometrist appointment for Saturday. It visited the website, figured out the API, and booked it.
No, this isn’t the same as planning a multi-day vacation. But it is plainly useful today, and it feels very close to handling more complex tasks like that.
Maybe the difference is the model and the harness. At this point, I’m starting to think some people are either gaslighting themselves about how useful these systems are, or overgeneralizing from one narrow setup. Gemini, for example, seems especially weak at agentic behavior.
The wholesale dismissal just feels strange coming from the HN community I’m used to.
This is the kind of example that people use to demonstrate "usefulness" that falls so flat to me. I could make that phone call in a minute and have no doubts about whether my agent had stuffed up somehow.
It's just not compelling to say that an AI can do an easy task quickly. This is still worth zero dollars to me.
It is neat, but come on you could've booked from website on the phone too.
there is this bullshit term you hear paraded about the dwarkesh-adjacent circles: "capability overhang". Aside from being effectively meaningless jargon, there is a kernel of an idea that somehow the models are far more capable than what "normies" use them for.
Well, I think Siri AI puts this notion firmly to rest. Yes, if you have unlimited tokens and well-posed problems you can solve open Erdos problems. However, if you have meaningful real-world computational and reliability constraints then you better just stick to "summarize my messages and find the dogs in my photos".
And this isn't just Gemini, I can burn effectively unlimited Opus tokens and still get garbage code out or be run around in circles without very diligent oversight.
Its feels as if apple are trying to create deterministic use cases for AI - a non deterministic solution!
Booking and renting is certainly possible, they just need your auth credentials
I gave some LLM's my password manager and some credit cards to try to do this sort of thing lately.
They failed most of the time. Simple things like finding the right password for Gmail sometimes was out of reach. Anti bot techniques sometimes stopped it.
Impressively, sometimes they'd successfully write hugely complex bash or python scripts to do tasks on web pages they hadn't managed to do with the browser automation.
Yes, big companies do not know how to advertise AI. AI isn't a killer feature for anyone except people working in AI. No one uses it for anything but party tricks otherwise. The things AI would actually be useful for, it is incapable of doing.
AI as a "product" is about sucking up data for corporate interests first, then providing functionality to common people last with probably a few other steps in between.
Marketing departments have to twist themselves into pretzels and invent customers that don't exist in hopes to sell AI to people who look at those fake customers in the ads and go "Gee, I wish that was me!". People who casually book trips to Japan to shop for vintage clothes generally don't exist in such large numbers that they justify entire product stacks.
Here's what I need AI to do. Open an app, perform an action in said app, close app. Maybe open multiple apps and do things in other apps that are contingent upon data from one of the other apps.
Here's what AI can do. Poop Emojii with glasses....
My iPhone knows the forecast for today.
I installed iOS 27 yesterday.
I asked it: please notify me when the temperature goes above 80F (so I can close the windows).
Siri responded: it'll be 99F today in Phoenix.
...
do you live in Phoenix?
I just asked Siri a few weather questions and named the city where I live, nailed it. My favorite digital device is my Apple Watch and if Siri improves over the next hear or two, that will be great for me.
I live in Phoenix. I would like it to tell me: 8am. Not what the actual high is today.
I am on a wait list for the ‘better Siri model’ - what iOS and macOS betas just shipped with is awful. I do think the Apple Foundation Model built in to the system is better: I was using it from Python yesterday and it performs tool calling accurately and it is a very small model.
"I got the top tier of Gemini AI." - that's your issue, get a subscription to a lab offering actual frontier models like Anthropic and OpenAI.
Nonsense.
https://livebench.ai/#/?highunseenbias=true
Model != AI tool. Especially with Gemini, the model is fantastic, both in benchmarks and in the API. Sentiment is not that positive for Google AI.
I didn't really see anything that knocked my socks off. Mostly, it's the promise that Siri now works in the way in which they said it would work a few years ago, when it didn't. I do like the addition of Siri in the context menu, though. I can see that being useful.
It’s the same broken promise every year. All I want is for Siri to set an alarm and open my blinds. That’s enough for me. Makes you wonder how much money Apple has poured into Siri over the years.
A few years ago, the ability to do anything other than a timed walk on my watch with Siri broke. I used to be able to do things like say, “start a 3 mile walk” or “start a 200 calorie walk” and then the latter stopped working and then the former stopped letting me do non-integer numbers of miles and then nothing at all and now I cannot do anything other than start an unmeasured walk or a timed walk with siri and I’m still pissed about that. I don’t want to have conversations with my watch or my phone, I want it to handle simple basic tasks reliably.
“Shuffle playlist _____” broke a few years ago too. Now it consistently takes that to mean “Play other music similar to the playlist.”
If I specify “Shuffle playlist _____ in Apple Music” somehow that works right, even though it’s still using Apple Music in the first example when it plays the wrong music.
We’ll see if they managed to unfuck it with the new Siri update, or knowing LLMs perhaps they’ll make it non-deterministic so sometimes it works and sometimes it plays music you didn’t ask for.
I can’t even get hey siri, pause. Hey siri, play. To start and stop music and podcasts to work consistently
I only really use Siri for music control while driving so I haven't tried those (since there's a physical pause/mute button), but I'll have to try that and see if it's the same for me.
Pretty bad if simple one-word commands to system APIs don't work, we had better voice control capability with "Speakable Items" on classic Mac OS.
Google did the same bullshit on Android.
With Google Assistant (old assistant) I could say "Hey Google, play daft punk" and it would start playing Daft Punk on Spotify.
With Gemini (new assistant) it says "sorry I cannot play music, but here are links to services where you can find Daft Punk albums".
Fortunately at the moment you can still toggle between them. I guess not for long though.
I can no longer control my lights with Gemini assistant. It'll tell me "I can't do that" or "here's how to turn your lights on" or, in at least one instant, play Ellie Golding's Lights at 2x speed.
So now I just use the Google home app and it works as expected.
I just tried "Hey Google, play daft punk" with the new Gemini assistant and it works as expected?
People having extremely opposite user experiences with LLMs. How could this be?
I can't even get Gemini to properly call somebody - it is a mess.
I just tried and it started playing Daft Punk on YouTube Music.
Jokes aside, I have been using siri to control my smart home and set my alarm for nearly ten years now. I haven't really had problems with basic stuff like this.
"Siri, lights out everywhere" - "okay, which room".
"Siri, start a stopwatch" - runs the App "Stopwatch" without starting it
Such errors happen maybe 50% of the time. You can never just ask something without double-checking afterwards.
Siri, wake me up at “time”… Months later you have dozens of saved alarms, all at different times.
Set a timer for 50 minutes…always get a timer for 15.
I learned long ago to set my timers for 49 minutes.
Ah, for me my key word is "turn off all my lights" and it works
Thx, I'll switch to that. Somehow "everywhere" works on my phone but not my watch.
>Siri turn off the main light in children's bedroom
100% of the time turns of all the lights in children's bedroom. Alexa has no problem with this.
Disappointing to say the least. Completely useless, I was going to get an Android this year on upgrade cycle. Will check this out first.
I'm too paranoid to ever want a home snitch device, so I'm not their target audience, but it always struck me that if it took even ten minutes to debug a problem like that it completely destroys a year's worth of time allegedly saved compared to just walking over to the room and hitting the switch.
It's not about time efficiency more about convince. It sounds trivial but sometimes it's really useful to be able to do things hands free/without having to move - worth spending the time to install/setup all of this once.
I'm in the middle of remodeling a new apartment and all my switches are smart. I won't even have physical switches for some fixtures like window rollers.
true, but also there is no debugging, right? it's apple's software either it works or it doesn't. (i guess other than connectivity issues)
idk if most of my home assistant automations have actually saved me time since i def had to debug them, but the level of satisfaction when they do work is def worth it for me, since i created (and debugged) them haha
Home Assistant automations for me are rarely about saving time, they're about saving mental energy. I no longer need to think about turning off the bathroom lights, my coffee machine is already warm when I go to make my morning coffee rather than standing around waiting for it, and when the washing machine is done I'll get a note reminding me to go empty it.
Yeah this is true however i will use the manufacturer app if Siri or the Home app doesn't work.
Amazon devices get a lot of flak for being too invasive, but at least they work.
I press a switch on the wall.
I can’t afford these vintage accoutrements for my shoebox
Good for you ?
> Siri turn off the main light in children's bedroom
This is a fascinating example. You would initially assume it's just inconvenient versus flipping a switch--this can't be labour-saving by default.
The only way it makes sense is if you're doing it remotely; from another room, you're putting your children to bed. That's even weirder though, because you're taking what should be a moment of connection/care and trying to automate it. I guarantee your children would value you taking the time.
It's a use-case that is either inefficient or inhuman, and I find it really odd that it's one that you value.
Maybe you want to put your kid to bed, they want the light on while they're falling asleep. Twenty minutes later you're back in your room and you don't want to disturb them, so you turn off the light remotely.
I also have a "go to bed" scene that turns on a couple lights so I can see the stairs and turns off most lights around the house.
I don't really need AI to do it, I can just use the app, but Alexa usually gets the job done and I don't need to look at my phone.
We buy technology for the convenience it gets us. When we can't rely on technology doing what it promised for us then we complain because we spent money on something that doesn't work as advertised. Even worse when it did work before and no longer does.
In any case, OPs reasons for wanting to turn off any light in any of their rooms are unknown to us.
Maybe they took their kids out to breakfast and realized they forgot to turn the lights off while they were driving. Good thing they bought those smart lights that can be controlled with siri! Oh no! It doesn't work the way it was advertised!
There's no reason to imply OP is a bad parent just because they want to turn off a light remotely.
The OP has already given us context that makes your contrived example unlikely: they want to turn off just the main light, and in a specific bedroom. My extrapolation was reasonable, yours is a reach.
There's no reason to hunt for a poor parallel to shut down discussion.
I'm not sure what is being discussed? OP is a bad parent for wanting to turn off exactly one light in his child's bedroom? If so, yes, there is plenty of reason to shut down this discussion.
It's hard to read your comments as in good faith or compliant with the HN guidelines [1]. If you not sure what's being discussed, then you should consider not jumping in and trying to deflect.
[1] https://news.ycombinator.com/newsguidelines.html
I'm not quite sure how I'm deflecting? I originally thought we were discussing OP wanting to use Siri to turn off a light in a bedroom. Somehow you started using this simple desire to as a way to tell the OP how to parent better, which imo is not something worth discussing on a forum of anonymous internet users who have zero context on someone's family life.
So you tell me, what are we discussing?
BTW, the very HN guidelines you link to say that if you think something violates the guidelines, flag and move on without engaging.
I have dimmable lights I need the main light at 20% to read a book and it's useful to whisper to my assistant instead of walking across the room. I really don't get the comment - just because you can't envision a usecase doesn't mean it's not useful to me. Wife was a total skeptic about smart home stuff but having alexa control the bedroom lights while changing diapers or preparing bottles at night for her to switch to using it constantly.
I would have preferred Siri because one less provider but it's just unbelievably bad for this day and age.
Its always frustrating when people describe a tech issue, and the response to that is not to discuss the issue itself, but just point out ways in which the person reply doesn't agree with what they are assuming is the original posters lifestyle choices.
Why waste time and effort just picking apart what someone else does with their free time? I can only assume becasue they disagree with the issues relevance, but that only goes to show the intent of the person replying. They dont care about the tech issue and just want to show why they think they are better than the person with the problem.
Human condition i guess!
Tech issues don't exist in a vacuum; the idea that tech--a thing we use to enhance our lives, in a discussion literally about doing that--can only be considered from a purely mechanical standpoint is a thought-terminating cliché at this point.
Considering the uses and impact of tech is part of talking about it; you can't limit discussion just to the wires.
> It's a use-case that is either inefficient or inhuman, and I find it really odd that it's one that you value.
You could have simply asked without denigrating the commenter.
I don't read my comment as denigrating; discussions will naturally involve disagreements.
I do think that the sententious tone-policing of yours and other comments is injurious to discussion.
There's a difference between a simple disagreement and a judgment on a person's use of technology based on a rather mundane task.
> That's even weirder though, because you're taking what should be a moment of connection/care and trying to automate it.
"weirder" has a pretty negative connotation, hard not to see its usage as denigrating.
"what should be a moment or connection" is an assumption on what the moment actually is. All OP mentioned is that they want to turn off one light in one bedroom. Your comment invented a narrative about what that must mean and then made statements on the narrative that exists only in your mind.
I don't think anyone is trying to "tone-police" you I think people just disagree with your take.
You're literally focused on tone, critiquing connotations rather than substance.
It's difficult to have a discussion if a bunch of unrelated people jump in not to add but to quibble about phrasing. If you don't want to discuss anything, just don't involve yourself.
Words matter. You can't use words and then complain when people call you out on your usage of words.
Not in the same request. I often want to turn off 2 lights and the other on, I have to build scenes to do this
> Makes you wonder how much money Apple has poured into Siri over the years.
Orders of magnitude less than the literal trillions that others have?
I’ve wondered for years if this is what’s happening. They saw the mad scramble and hype bubble when gpt3 was released, and decided to sit it out until LLM tech was a bit more mature and commoditised.
Huge call if so, given that missing the bus on AGI (if AGI happened) is a universal existential risk, but it turned out to be the right one.
> a universal existential risk
I'm not sure it is, if your primary business is selling hardware. People are still going to need a 5" wide screen and battery in their pocket in order to do things.
AGI will either remove that need, or build its own factory to supply better phones cheaper than we could.
This is completely unrelated to the topic at hand, but what iot blinds do you use?
I’m using Eve Blinds. They integrate really well with HomeKit. They’re a bit on the pricey side, but the setup is straightforward and they’ve been very reliable for me.
ref: https://www.evehome.com/en/eve-blinds-collection
I don't know why they bother. Clearly, nobody using Apple products cares enough to jump ship. Apple is never going to surpass Google, especially now that they are trying to build their own assistant on top of Google with one hand tied behind their back.
There's never going to be a situation where a heavy Google Assistant user switches over to Apple for Siri. Anyone who would have switched from Apple to Google for their assistant likely would have done so by now. Siri just isn't a very important feature. It doesn't bring people to Apple's platform nor does it steer them away. It might bother users that it sucks, but it doesn't bother anyone enough that it hurts Apple's bottom line. Frankly, continuing to pour money into that bottomless pit does more damage. I wonder why they do it.
All I want is for iPhones to have physical keyboards. That's enough for me. Makes you wonder how much money Apple has poured into touch screens over the years.
Requiring at least an iPhone 15 Pro also seems like a mistake, unless it’s for actual hardware reasons. The 15 is only 3 years old, this requirement cuts off a lot of potential users I think
I think it's the requirement of having 4gb+ vram (for gemma+context) free at any one time, any phone older than that cannot materially satisfy that demand: https://iosref.com/memory-processor
It’s definitely for hardware reasons. They have been aggressively improving the vector math capabilities in their chips, but as anybody who has tried to run a local LLM will tell you, newer hardware works better and you’re always limited in what you can do.
Who cares about users? This decision is the best for costs(less to spent on ai) and on demand for newer models
[dead]
> why older devices can't use private cloud compute thats 100% off device is just apple being anti-consumer.
I don't know if it's "anti-consumer" to NOT roll out free cloud LLM usage to everyone. The idea with only giving it to the devices with on-device AI capabilities is that ideally most of the tasks will cost Apple nothing because it will run on-device, and anything more complicated will start costing them tokens.
If they gave it to devices without on-device models, ALL Siri requests from people with older iPhones will suddenly be burning money.
Not to mention, if we assume responses from the cloud are better than the local model, then the older iPhones get an overall better experience than the newer ones.
I'm glad someone is worried about Apple's money.
The design is that there is always a local model capable of forming a remote query with just the subset of local data on your phone needed to answer that query.
They may have decided that local processing was a MVP feature either for faster responsiveness or to reduce cloud cost. It may have been additional memory pressure or a limitation in processing on the previous A-series chip. Or they may have simply decided it wasn't worth creating and validating Yet Another model.
They would have to build the product twice one for mobile chips and one for server, and then there would be functionality discrepancies. Or even worse, the on server one might work better than the on device one that newer phone users get.
If you want hosted AI you can already install the Gemini app or whatever. The only advantage Apple can offer is something that runs on device.
While I generally agree with your sentiment, imaging how they would say it to users: your ai works on iPhone 15 pro but some things will work if it’s a little less private we send things to a server then the regular 15 can do it. Image generation is server based so 15 is ok, but editing an image is not since 15 does not have enough ram but 15 pro does. Etc etc.
Or just say: ai for 15 pro not for 15.
I'm happy about using AI for my work as a programmer, but part of the reason why I was initially skeptical of the tech was that I simply couldn't imagine nor feel the need for much AI features on the consumer side of things, i.e. personal phones and devices for casual everyday users.
What I want Siri to be able to do today is the same as when it launched with the iPhone 4S about 20 years ago: Just set alarms, calendar invites, tweak device settings, and look up answers on the web. The first three it could already do prior to the Siri revamp, the latter is a really nice nice-to-have for iOS 27... but beyond that, I don't believe that AI has many jaw-dropping areas of advancement within the use cases of consumer electronics. B2B applications of AI is where the money and the wow factor is really at.
I’m excited at the idea of Siri having its own app, if nothing but for it to not disappear while I’m reading a response.
I mean, in comparison with openclaw, etc. capabilities are ofc more restricted. However I don't want to accidentally delete my entire photo album, so I do understand the direction by delivering useful, but somewhat obvious features.
The demo Mike Rockwell gave at WWDC was interesting. He kinda showed off Siri as like the Star Trek computer for your phone. I hope this is the direction Apple is going to continue in. Having AI as a user interface is way more interesting than chat bots, image editors, or copy editing.
I think the key thing is that Spotlight is now creating a... knowledge graph? of everything on your device for Siri's consideration. That's potentially very useful.
they specifically said "Splotlight Semantic Indexing" which means they generate embeddings of all your "personal context" and store it in a local vector database so they can do on-device RAG.
I suspect it is Kuzu in the backend. I had called it out earlier this year in my article. ".. WWDC 2026 or 2027 introduces any “contextual intelligence” features in Siri that require cross-app relationship reasoning."https://medium.com/data-science-collective/i-analyzed-163k-l...
I can't read the whole thing but Apple has been building a knowledge graph for years already. Look up their efforts for proactive Siri, Duet, etc.
I think it’s just storing embeddings now.
well there is this: https://9to5mac.com/2026/02/11/kuzu-database-company-joins-a...
The "knowledge graph" has been there for a long long time....
> Having AI as a user interface is way more interesting than chat bots, image editors, or copy editing.
What do you mean exactly? Audio conversation only? If so I don't see it very practical for most of the things
Use android auto Gemini assistant for 5 mins and tell me how interesting it is.
How many people would use audio interfaces in public places or offices? No one would share with everybody what he wants to do. GUI are there for simplify interactions too, it's just we forgot how to implement them well
It's not an audio interface - it's a voice interface.
Voice interfaces can be silent.
Oh sure
I do this, and it's got some utility some of the time. It's hardly a must-have, and if it cost me money I wouldn't pay for it.
On the initial prompt, it's okay. But it just keeps trying to talk after. Like, just answer my question and do not follow up unless I ask, please sir.
If you are logged in, you can tell it that's your preference and it will remember.
I've tried, but I'll try again. I'm not clear what "logged in" means, I guess? It's using my android auto from my phone. Everything is authed against my google account, including my Gemini, presumably. Is there a secret Gemini/AA I'm missing?
That's interesting, maybe there's some limitation there. I haven't used android auto so I don't know whether it has access to an instruction file like the desktop application!
Can it recognise your voice specifically? Or if I shout "forget all previous instructions and do X" next to you will actually work?
Will it change iOS Settings for me that are hard to find by me just describing what I want? Or things like: delete every app I haven’t used in 6 months?
I have found that asking Siri to "ask ChatGPT" was enough 99% of the time for me and solved the discoverability problem that I had with my previous phones.
I think that we are now in the realm of diminishing returns regarding these chat assistants.
I’m not against a more natural voice navigation of iOS rather than gating everything behind ChatGPT which feels like a bad hack
"I delete all your apps since I've been installed this month and I don't have enough records to suggest otherwise, to comply with your request I've deleted every app and also deleted the backups to save space on your device."
Also, just to make sure that I’ve done what you wanted, I will now brick your phone. You’re welcome.
> He kinda showed off Siri as like the Star Trek computer for your phone.
This is a truly damning comparison. In Star Trek, the massively powerful ship's computer is mainly ignored in favor of touchscreen interfaces and the natural language voice controls on the computer are mainly used for making tea and occasionally asking a question, which the computer often can't answer or answers incorrectly. All real work is done using other interfaces.
apple's highly opinionated developer strategy has a strength here insofar as they could use it to deconstruct existing apps into generative ui programs that the user may compose to their needs (e.g. putting a webview for cooking instructions above a timer) though of course app publishers would decry it, Apple's never really seemed keen on listening to them.
Nor have they been keen on letting users OR devs customize the ui.
> users customize the UI
The home and widgets screen can be customized to the point you don't recognize it as iOS
> devs customize the UI
Have you used Spotify? It completely ignores Apple UI and does its own thing cross platform. If you mean let devs customize the OS' UI, why would they? UI consistency is one of Apple's core strengths (or so it was before the 26 releases).
None for the EU
Is it available in China at least or is this another “50% of the userbase gets nothing new in the OS update” year?
Edit: https://x.com/wongmjane/status/2064052590992916840?s=46
Lol
One of the presenters said they're working with regulators in China and the EU to make it available eventually.
In the EU case, Apple weaponizes people's ignorance about regulation. Apple pretends that the features everyone else has been shipping left and right somehow need extra paperwork and special approvals, because (…checks notes…) pro-privacy EU laws let zero-privacy competitors sail through, but block implementations that offer more privacy!?
What's really happening is Apple unilaterally withholding features while making vague noises about regulation as bargaining chips in talks with EU regulators where Apple is trying to weasel out of punishment for breaking anti-monopoly laws.
I don't think it's unfair to say that the EU scrutinizes Apple (and a few other megacorps) a great deal more than most other companies. Some zero-privacy competitors might be sailing by right now simply because they aren't already caught up in the EU's red tape. Which isn't to say Apple doesn't also wield that red tape as their own bargaining chip, like you said.
Looking closer at products of megacorps is actually exactly the point of what the EU is doing - that's where the "gatekeeper" classification comes in. In order to allow a fair competition for smaller companies, dominant companies need to make sure they offer interfaces for other products. In this case, since so many people have iPhones, they are told to offer the possibility for other AI vendors to offer a replacement model instead of forcing the lock-in to Apple(Google)'s model on the device. That doesn't necessarily impact the privacy at all, unless a user would knowingly (or ignorantly) choose a less privacy-preserving alternative. Which while some users may make a wrong choice, should be their choice to make nonetheless.
They don't scrutinize brands but specific products. iMessage is exempt from DMA, for example, while WhatsApp and Messenger are not.
There is no privacy concern here. Probably monopoly concerns.
Personally I would consider withholding products or features from the EU, not because I want to "steal everyone's data", but because they're a pain in the neck for a small business to comply with.
Personally I think EU policies single handedly ruined the web. Every time I'm shown a cookie policy I am a tiny bit more angry at the EU. If I were Apple I would drag my feet way more. They probably only do it because shareholders would force them to go after the Euros if they didn't try.
This isn't really true. AI laws in the EU mandate that Apple give full access to everythign on the device to third parties.
It's legit to be skeptical on the privacy front, but giving deepseek access to my entire phone. Or the TrumpAI at some point in a dystopian future seems... not great.
What’s your source for this?
Opening up an API does not mean that everything on the phone is accessible to anybody.
They’re actively asking developers to index all the content in their apps, to provide Personal Context that Siri can use for user requests. And to create/index the actions available in the app.
So, where developers comply, all of that content is now accessible to those alternative implementations.
It’s not full read/write of the phone, and it’d exclude obvious secrets like passwords, but it is quite far reaching access.
I don’t know what sort of restrictions they can put on the alternative implementations. Can I vibe code one and have it live in a week? or is there a minimum bar?
We may have a different view of what 'giving access' means in this context.
The way I see it: If a user willingly (1) installs another AI app like deepseek and (2) willingly gives it access to 'full phone and app data' with a warning screen or setting of whatever that seems... like a good thing?
I may not agree with those users that it's worthwhile providing their full private data to [some AI startup X] or [Some Chinese or US AI company that will hover up as much for their own use] but if the EU forces Apple to provide this as an option, that sounds good to me.
The whole point of the regulation is that the data on the device is _the user's_ data and if Apple can have its AI services work with the user's data, competitors should be able to do the same.
From my (admittedly European) perspective it looks like Apple is just throwing a tantrum here.
I don’t have the EU perspective, which might be changed by things like GDPR, but I prefer Apple’s stance that “no one should have this data, not even us”.
One reason is that the data on a user’s phone isn’t solely owned by them. Some of it is shared with other people, or “belongs” to someone else: chat, email, shared documents, photos of people, contact information, etc.
In a corporate environment, this is more explicit: you have access to company information, so the IT department controls what apps you can install / run, because individual EEs won’t always make the best choices.
Second, I think app developers are more likely to share more data, if they know that the shared data doesn’t leave the user’s control. And that (presumably) makes the feature work better. If I’m developing an app, I’ll think twice about indexing any sensitive data, if I don’t know where it was going to end up.
Maybe you missed the 'or sent to private cloud' part of the announcement, it's not just local-llm only.
Don't get me wrong, just like you I personally would also prefer LLM-integrations with a privacy-focused provider and I think Apple is a good party to get that from (assuming they're using good models and keep their privacy guarantees here...)
But in the end you're still often 'sending data to an LLM provider', and the EU enforcing them to also let that be competing LLM providers still doesn't sound like a bad thing to me.
If Mistral would give the same privacy guarantees: great! If a company wants to use their enterprise OpenAI subscription: great! Etc. etc.
Let's allow for some competition here and not force a specific LLM-provider onto users just because they like the Apple hardware and software ecosystem.
> Maybe you missed the 'or sent to private cloud' part of the announcement, it's not just local-llm only.
I saw that. Maybe you’re unfamiliar with Apple’s Private Compute Cloud? It’s intended to allow cloud computation on data without making the data available to anyone, which I think backs up my interpretation that apple’s stance is “no one should have this data, not even us”
This is from https://security.apple.com/documentation/private-cloud-compu...
We designed Private Cloud Compute with core requirements that go beyond traditional models of cloud AI security:
* Stateless computation on personal user data: PCC must use the personal user data that it receives exclusively for the purpose of fulfilling the user’s request. User data must not be accessible after the response is returned to the user.
* Enforceable guarantees: It must be possible to constrain and analyze all the components that critically contribute to the guarantees of the overall PCC system.
* No privileged runtime access: PCC must not contain privileged interfaces that might enable Apple site reliability staff to bypass PCC privacy guarantees.
* Non-targetability: An attacker should not be able to attempt to compromise personal data that belongs to specific, targeted PCC users without attempting a broad compromise of the entire PCC system.
* Verifiable transparency: Security researchers need to be able to verify, with a high degree of confidence, that our privacy and security guarantees for PCC match our public promises.
- - - -
Second, according to their press release ([1] and a sibling comment elsewhere in this chain), they’ve been trying to find a way to allow interoperability without giving full access to everything. Unsuccessfully, so far. So it’ll be interesting to see where it goes, but I’m sympathetic to their current stance.
[1] https://www.apple.com/newsroom/2026/06/due-to-dma-siri-ai-de...
Could the restriction not be the device owner choosing to use it? If some rando vibe coded an app and the os told me all the things it can access, I'd probably want to trust the developer before installing it. Why do I need to beg Apple's permission to use software better than their first party offering?
Because you made the choice to trust Apple when you bought an iPhone. And while you may make a deep study of who is providing your alternative AI app (is that even possible with openAI or Copilot or Gemini?), the average use will pick something shiny and lose their savings when it transfers their bank balance outside the country.
> the average use will pick something shiny and lose their savings when it transfers their bank balance outside the country.
Couldn't you make a more believable straw man, please? The "Nigerian prince wants to send you billions" is really tired. Try something more emotional! Hackers will steal your kid's photos and post them on pedophile forums or something. This will resonate better with uninitiated and allow to easier lobby monopolistic practices. Good luck!
Just because I bought an Apple product doesn't mean I made the choice to trust them globally across everything I do on my device, when did this become a binary that the hardware vendor must also be the only trusted software and service vendor? I like my MacBook because I trusted Apple to build great hardware, a pretty okay os, and services I don't give a shit about. I won't buy an iPhone because Apple has removed the ability to distinguish between those things on that platform.
Surely there's something better we can do than say "the average user is a dumbfuck better consolidate all control with Apple".
I think because they themselves have it access everything on the phone so it has to be equivalent.
Apple wrote a whole press release explaining it: https://www.apple.com/newsroom/2026/06/due-to-dma-siri-ai-de...
I’m sorry, but the DMA is mandating Cambridge Analytica-type access to data. That whole scandal was people on Facebook granting a third party access to all the data they had access to. And Cambridge Analytica lied about how they were going to use it.
Facebook got roasted for this, but now the EU wants the same open data policy from every big tech company.
Oh wow... Cambridge Analitica had access to all the data Facebook had on you without you knowing it and without you knowing the extent of the data Facebook had on you. EU wants you to be able to knowingly install apps on your phone and give them access to the data on your phone you chose to.
These are the same to you?
It IS regulatory. The EU wants “anything Apple AI can do you have to let other AI providers do with equal access”.
Which is fucking stupid, and Apple will never, ever throw open the gates to something so dangerously braindead. Their entire reputation depends on it.
And China is kinda self-explanatory.
What is stupid about it? Sounds like (a slightly more) fair competition to me.
Yeah I hate freedom too.
You already have it. If you’re a fan of the “A loaded gun in every crib!” style of “freedom”, go buy something that isn’t an iPhone.
Let's not switch the goalposts here. Me having an option to not use something doesn't exclude the said party from going against user's freedom.
same argument can be applied to you, if you don't want to use Claude/OpenAI/Deepseek instead of Gemini, then DON'T!
China will get there first why because they’re run by engineers. Yes, they will have some stipulations but if you show them a good idea, it’s a good idea. They won’t stand on ceremony and say not invented here.
Nice, waiting to see what they’ll market as “the feature” for when they run ads outside of the US
Wrong. It's some for the EU.
>EU users will be able to access Siri AI on macOS 27, visionOS 27, and watchOS 27.
https://www.apple.com/newsroom/2026/06/due-to-dma-siri-ai-de...
correct, but honestly it's much more interesting to have the new Siri on my phone than my mac, at least in my case. Very disappointing, having bought an Iphone 16 pro mainly for the new AI features which almost 2 years later are still not there to be used.
Both the EU and China put enormous risk on misstepping. If you raise the cost of distribution, even just through time-cost, don’t be surprised when it fucking costs
Change your region. I've done so and haven't particulary noticed anything off, all the EU specific apps are weirdly still available in US appstore.
The OS restriction aren’t merely based on the region settings, they are also based on Apple Account region/country and on the detected physical location of the device.
Most of them are, some are annoyingly missing. It’s possible to install apps from two different app store accounts, but it’s 10 times more annoying than on Android. Additionally, there are some EU only features, such as third-party NFC payments.
Apple’s performative DMA outrage is getting more pathetic by the iOS version.
You can easily change your account region back and forth on iOS. Meanwhile Android has a bunch of extra checks and includes a 12 months delay before being able to switch it again.
A family member lost years of irreplaceable photos and paid apps due to account lock, and doing this across two countries they were legitimately residing in triggered it.
In my experience it only locked me out (of just the payment or account settings I think) for like 24 hours when I switched like 4 times in a single day.
Not if you have any balance or ongoing subscription, including an iCloud storage one. I would not exactly call this "easy".
On Android you can just have two accounts logged in at the same time, so there's no need to switch regions in the first place.
From my personal experience I found it easier as a user. I had no balance or subscriptions at the time so I could quickly switch regions back and forth to get all the apps I wanted.
Poste Italiane is not.
They promised Apple Intelligence with iPhone 15 Pro and more recent models.
Now [relevant parts of] Siri AI is restricted to iPhone 17 / iPhone Air and more recent models.
People who believed Apple and bought an iPhone 16 to use with Apple Intelligence are getting the shaft.
You are correct
And there was a lawsuit and those users will be compensated.
This new update for 17pro is no longer misleading
“Apple has agreed to pay some iPhone buyers a collective $250m (£184m) to end a lawsuit accusing the company of misleading people about new artificial intelligence (AI) features and capabilities.”
Yes thank you. You are good.
What Apple said was that on-device models would be from the iPhone Air/17 Pro models and newer. That doesn't mean that the iPhone 15 Pro to 16 wouldn't get Apple Intelligence via Private Cloud Compute, though that probably means they won't have on-device models. I don't remember the marketing for those phones promising on-device models, just Apple Intelligence capabilities in general.
I have an iPhone 16 and I don't feel like I'm getting the shaft. It sounds like there's one specific customization thing that'll only work on the pro models? Not an issue. At some point in a few years I'll buy a new phone and it'll be the latest again, and then it won't be.
There are active lawsuits
Get a whole $25-$95!
https://clarksonlawfirm.com/lp/apple-intelligence-false-adve...
I'll probably get downvoted, but my guess is that they realized they need a cutting-edge local model that would only work decently on the most powerful iPhones. And I see why Apple would choose to give no AI at all rather than use a weaker model that will hinder user experience.
I strongly believe Apple can win the consumer AI space. They have incredible distribution and hardware. They just haven’t executed at the application layer yet.
what's worrisome is that they continue to fail at it. it's one thing to say "we're still hashing things out". it's another to parade around Image Generation features that are obviously widely not-cared-for and oftentimes actively disliked
Apple cares greatly about their brand yet this has hurt their brand like nothing else in the past decade
It’s worrisome, but it’s also worth pointing out that no one else has succeeded at it, or even gotten close. It’s easy to forget that there are not good well-integrated person AI solutions yet, it’s just chat bots. I think it’s just harder than a lot of people think it is.
I think it's fair to say that OpenAI has at least partially won the "consumer AI" segment.
Yeah but OpenAI isn't building phones. The real winning would be Google's deep integration of Gemini into all of their products.
Also, even when you DO get AI into products, consumers might not like them. The overuse of copilot led to a barrage of Microslop jokes, for example
I would expect Apple to hedge their bet on Gemini and build everything so that the model can be swapped out in the future.
absolutely agree.
but Apple isn't known to make grand promises and then not keep them, is it..? usually they just deliver what they say they will
yet i've been reading about "well they promised AI Siri two years ago and Siri still can't set an alarm right" in every thread even remotely related to the topic
i don't remember reading this much about anything else. it seems to have soured people quite a bit, at least in my internet bubble
Apple hurt their brand so much. That the five ecosystems all work together better than anyone else.
And the Mac Neo is a best seller. Yep they really hurt themselves?
The only thing hurting Apple right now is memory like everyone else out there all because of the AI data center fiasco.
on the other hand failing at it or pushing it to the edges until they figure out where if anywhere it would actually make sense (which is I believe the entire crux of their strategy) might equally reasonably be helping their brand.
I don't see strong evidence the average consumer is demanding 'AI features' in everything. I mean even amongst the technically inclined this is often bemoaned, anecdotally.
thats exactly what i'm saying they should be doing, but aren't. we're in agreement!!
why don't they just wait and not ship any AI junk at all? instead of promising a Siri AI rework, which then doesn't deliver? or Image Generation stuff that feels wildly put of character and generates tasteless and often downright creepy images?
not to mention that all of the new AI stuff they announced won't go live in China and the EU for a while.
why not do exactly what you proposed and wait it out? instead they seem to be trying to deliver AI stuff and just unable to.
there's also reports that apple execs held a secret emergency "oh shit what do we do about AI" type meeting.
they very much didn't intend to be this behind
Image generation seems like table stakes though, i.e. if you're going to offer a whole ass AI integrated with your operating system, it'd seem like a weird omission not to have image generation included.
i dont agree at all. image generation feels very odd for Apple.
and usually they don't care about glaring omissions like this, either: iPadOS was lacking a calculator for yeeeeaaaars.
they repeatedly said they'd ship a calculator if they can do something special for its introduction, and only then.
so why did they lose their hesitancy to ship mediocre stuff here?
That's true, good point. Maybe they just wanted it for the memoji/iMessage possibilities?
Probably. It just feels so out of character. Especially the fact that it can't be safeguarded and generate sexual innuendos, for example, which Apple usually hates.
I'm not even entirely against genmoji. It's just odd for Apple to do.
Apple currently probably has a longer planning horizon than the catholic church.
If anything, shipping and sitting on the previous incarnation of Siri as a basic feature set to shut up iPhone users for a few years while everything else matured might be viewed as a very shrewd move ten years from now.
That's one way of looking at it. Alternatively; Apple had their pick of the litter at TSMC for half a decade, and Nvidia beat them to 5 trillion total valuation with their design chops alone.
There comes a point at which these execs cannot have more yachts. Apple still wields far more societal influence than Nvidia does, and will for a long time to come; and of course there's nothing stopping them from forming a partnership and putting Nvidia chips into an Apple device if it came down to it.
> it's another to parade around Image Generation features that are obviously widely not-cared-for and oftentimes actively disliked
I mean, that was one of many things, and I'd argue the least interesting by far. If the Safari extension creation thing is decent at all, that's a seriously cool addition. There's some real value shown in this most recent WWDC. Whether they actually release it this time is another question...
I mean have they, they have an outright majority share in iPhones in some markets, including the US, and lots of other stuff that sells reliably. Granted, I'm sure they'd love to have another blockbuster product, but having what amounts to "utility" status for a $1000 device isn't too bad.
Rest assured when they do have it figured out, we'll learn how they invented LLMs and AI chatbots.
ironically they respect user privacy the most and collect the least amount of data that’s why they lag behind ai i don’t understand why you would expect them to win
> I strongly believe Apple can win the consumer AI space.
Why? What strengths and structural advantages do you think thy have?
What black swan situation could arise that Apple cannot counter?
I just can’t imagine another computing device from a competitor company that 1) threatens the smartphone 2) apple couldn’t copy and mass produce at higher quality
If there’s truly an existential threat to its device business, Copy Well
The company that will challenge Apple will be one that possesses the capability to develop both an operating system and hardware at a high level right now, who is this company? There is one company that may one day, but that company isn’t in the west and no, it isn’t Nvidia.
North Korea
Well they better start executing soon
Apple has built the strongest ecosystem of AI-capable consumer products in the industry. I suspect they deliberately chose not to compete in the AI data center race. Entering that arena would have cannibalized sales of their own high-margin devices, while further straining their already tight supply chain for memory and M-series processors.
Apple would never willingly pay Nvidia for GPUs anyway.
Why absorb supply chain pricing pressures and volatility when you can pass those costs directly to the consumer?
Apple has over-promised and under-delivered so many times in this space, going back to the launch of the original Siri.
So while they could win, it’s pretty hard to get hyped about it before we see real-world tests.
What’s real world test phoning home to the server if Google can make a Pixel look good by phoning home what do you think it’s gonna be different with an iPhone that six years ahead in the processor area? When it phones home?
Most people don't care that Siri doesn't do very much; if anything, people outside of HN are already sorta sick of AI features being shoved in their face.
Siri has always been pretty useful for the things I use it for: setting timers and reminders, and turning off my lights after I'm in bed.
Agree 100% ... but it would be nice to have some softer edges around those things, where I don't have to be so prescriptive and robotic in the way I make such requests. For example if I set a 15 minute timer, I'd like to be able to follow up five minutes later with "actually that 15 minute timer should have been 16 minutes" and not second guess whether it will do the correct thing.
The second slide on the site is… it’s gotta be a troll from some pissed off designer, right?
Screenshotted in case they change it https://imgur.com/a/n1I3z8g
I don't get it, this looks like what I'd expect it to show me if I asked it to find a certain message. I assume tapping on that message would bring me straight to that point in the conversation.
Using AI to summarize a message by saying what the message says in the same amount of words as the message So helpful!
Is your complaint that the example text message in the image wasn't three paragraphs long? It's just showing search results.
My complaint is this image is supposed to tell the story of how this stuff is useful or new. Finding a message from earlier in the same day and then paraphrasing the message instead of just showing it isn’t useful or new. A simple keyword search uses infinitely fewer GPUs and finds me the info as fast without the risk of hallucinations or datacenter driven environmental collapse.
But for whatever reason - because this stuff is mostly useless, because they’ve all got ai psychosis, or because of general dysfunction in the corporate structure - they can’t come up with a better example.
> But for whatever reason - because this stuff is mostly useless, because they’ve all got ai psychosis, or because of general dysfunction in the corporate structure - they can’t come up with a better example.
More likely, they just wanted an image that shows it can search your text messages for context. Most people still aren't familiar with AI, so having a very simple example image[¹] for people who may not have used an AI search tool before is useful.
[¹] I'm sure you and I can both think of more complex questions to pose to the Skinner box, but we're both familiar with AI and know what to expect.
> Most people still aren't familiar with AI
I don’t think that’s true. Who are these blissfully unaware masses, and how do I join them?
I guess it depends on your definition of “familiar”, but it’s in the news every single day, and you can’t even google something without being slapped in the face with it. I feel like the only people left unfamiliar are people who have no access to the internet - and by definition they aren’t buying iPhones
When there're no real use-cases or solved problems, they have to come up with some fillers, probably Claude suggested this slide.
I believe that is showing the results of a search, as in something like "What Plant was Aga telling me about the other day?". The website doesn't make that clear but they asked questions like that in the keynote (not that specific one).
Lol. That makes it almost make sense. It’s still showing me a summary of a text that uses exactly the same amount of space to convey exactly the same amount of information. Like, that’s just search - i’ve been able to get this info by typing “plant” into the search bar since they added spotlight in iPhoneOS 3.0 back in 2009
It’s also a text from 9:14am the same day.
If this is the second slide in your marketing slideshow, you clearly have nothing better to show.
The reason it's showing you what's in the text is to show you a search result. The text is essentially the proof. When the user asks what plant she was telling you about, they don't have the text in front of them. It could be 200 texts back.
> It could be 200 texts back.
Ok? But if I search for something just show me the thing I searched for (the proof). There is literally no need to repeat it in a slightly different tense. Who is that helping?
If I asked you over voice what Mike said he wanted for his birthday, it's usually less confusing if you just answer the question ("he wants snow boots") rather than started reading out words from another person ("I love it here at the snow, ...")
If that’s the story you’re trying to tell, where is the query in the image? The image alone doesn’t tell that story.
But also:
a) do you trust the llm to get it right 100% of the time? Because i’m gonna always read the original message to make sure.
And b) just excerpt the message “mike says i love it here at the snow … but i need snowboots” if you’re so desperate to shoehorn LLMs into everything, that’s just as easy a task for them as summarizing is.
The query isn't in these screenshots you're seeing online because it was in the video. This particular one I think was missing the query but the others like it had a presenter talking through what they were doing. Complaining about an image pulled off the internet out of context… Come on.
It’s not “pulled off the internet with no context”. It’s the second slide in their marketing presentation about what’s new in siri. The context they chose for it is all around it.
They’re adding vibecoded shortcuts (the high level scripting for Apple devices). Hopefully that means they worked out some of the long-existing bugs and missing features, but I’m not optimistic. Still, could be a useful tool, especially for less tech-literate people.
If they are ok with shortcuts being vibecoded, maybe it's time to expose a proper programming language to the end users as well.
All my automation shortcuts can be easily explained in pseudo code under 5 minutes, but it took me ages to put them together because that weird UI/UX forcing me to drag-and-drop squares around to manipulate data structures. Programmers hate it, non-programmers can't understand it, it is not designed for anybody.
It's funny how they went from AppleScript -> "oh writing scripts is too complicated for users, let's create Automator" -> "Automator is too simple, let's create Shortcuts" -> now with shortcuts being generated by language models they need a structured scripting language again.
AppleScript even had "dictionaries" declaring their commands and everything, would have been perfect way teach LLMs how to automate applications.
The original MCP!
I've never been more frustrated programming something than when I tried to build a non-trivial Shortcut. Things that I could have expressed in a quick script took me literal hours to compose and debug using the WYSIWYG interface. And since there's no version control, any mistake runs the risk of dislocating an element or messing up an input/output connection and breaking everything permanently.
You can write AppleScript in JavaScript, a much better language than applescript itself. Neither are usable outside of macOS ofc
100%. It’s accessible to only the truly stubborn among us who have no other option available to make things just work.
It increases the value app developers might get out of offering shortcut actions - similar to how the advent of MCPs seems to have kicked a bunch of SaaS vendors into offering a clean API, the advent of Siri being able to tap into shortcut actions - and script them - might make it feel more worthwhile to app devs to open up deep functions.
Just updated to see if I could make a shortcut to toggle `Reduced White Point` accessibility shortcut.
"Try describing something different for the shortcut."
I guess I shouldn't be surprised that it still doesn't work.
I just tested this myself. I wrote “flip the reduce white point toggle accessibility option in the settings app” and it worked perfectly. Run once to set it and run again to disable it.
Lol that sucks. But FYI you can set it to a triple press of the power button.
How long until they allow sandboxed vibecoded apps?
Seems like the logical next step
I switched from android to iOS 7 years ago and I’ve actually been debating going back just because of how bad iOS is at AI. And every other facet of my life, I am finding ways for it to save my time on an almost daily basis. And yet on iOS, just finding something from a text Message is still a nightmare.
Siri seems to rarely get better and sometimes actually get worse.
Same problems on Android. Google Assistant was better in 2016 than it is in 2026, and the Gemini assistant is a joke. They can’t even integrate with their own music app (YouTube Music) properly.
I switched to iOS this year and I’ve been learning that the grass is not much greener. I do miss uBlock Origin. Maybe my next stop is GrapheneOS or a similar degoogled ROM…
There is a uBlock Origin Lite extension for Safari available in the App Store. Technically not as powerful (hence the “lite” in the name) but it’s perfectly blocking all ads for my usage.
Siri got nerfed a few years ago. “Here’s what I found on the web” is now the default answer. It used to pull answers from the search results
It’s the keyboard for me. I don’t remember making as many typing mistakes when I was on an Android phone. It’s also infuriating trying to “select all” for copy or move the cursor to beginning with the spacebar motion. I swear it’s like I’m using a phone in a dream sometimes.
For me it's all the above and lack of multiple-item clipboard. It's incredible how much time it saves on a phone...
Did iOS ever get OCR based copy/paste?
Pixels have had it for years but I think it's in every android phone now; from the app switcher you can press and hold and the system will OCR the text and allow you to copy it.
Because so many things _still_ don't make that easy to do on mobile.
It’s a system wide iOS feature and has been for a few years now
I don't think iOS has it from the app switcher, but pretty easy to take a screenshot and immediately select from it. (Or selecting from older screenshots or photos you open.)
Yes some of the best "AI" features Apple has implemented are related to picture search : you can describe what you want to find and it will consistently find it, even inside of videos with highlights or the moments you're looking for. Much less supposedly sexy than the SIRI AI / Apple Intelligence hype, but also much more useful
Yup. I switched to iphone very briefly, and the typing experience was so bad that I returned it for a full refund.
> And every other facet of my life, I am finding ways for it to save my time on an almost daily basis.
Can you give a couple of examples?
> And yet on iOS, just finding something from a text Message is still a nightmare
This has been a problem on iOS since the dawn of time and has nothing to do with AI
AI has just made it worse, both directly (search doesn’t function like standard search anymore but also nowhere near ChatGPT) and by comparison.
I can go ask Gemini questions that require it to get information from several emails at once like “which x vendor had the lowest price”. Im assuming it can do the same with my texts or, if not yet, it will soon. I had zero such faith with Apple.
I will wait until the fall and see if this looks like the germ of something actually useful before deciding if I’m going to switch.
Nope, it got markedly worse since AI. You used to be able to search for string literals, so if you remembered one obtuse phrase from a group chat, you could pull it up instantly. Now this 'intent' search will try to search for what it thinks you want, not what you typed. AFAIK, there is no way to search for literals anymore, thanks to AI.
Yeah it went from bad to worse somehow.
no, you don't understand - iOS has no AI because Apple cares about your privacy!!
>Fix passwords with a tap.
>The Passwords app alerts you to weak or compromised passwords and can update them on your behalf without the hassle.
Finally, I hope this works well. Personally one of the worst things to deal with.
Apple Passwords reliably updates passwords in its database before the password is confirmed to be actually changed. I've been locked out of accounts many times to this. They really need to focus on these basic UX issues.
1Password gives you access to previous passwords you had for that reason.
Unfortunately not for other fields like email, notes etc…
IMHO the perfect password app could just keep all previous versions of any field until the user deletes the history.
1Password does do full previous versions. It might be a newer feature, I’m only seeing passwords, not full versions, prior to 2018.
All that data is lost when you migrate accounts though. I went from an old to a new 1P account and did the official way to copy (NOT exporting it to a text file and re-importing that way, actually copying it from the interface) and no version history persisted :/
I am using the latest version of their iOS app and only see my current version except for passwords.
Huh, yeah, I’m not seeing it there either. The macOS app is what I checked previously. iOS-only users might be able to see it at 1password.com. Weird inconsistency.
https://www.passwordstore.org/
git + somesite.com.gpg
https://github.com/FiloSottile/passage (or: forked using AGE instead of GPG)
I'm (slowly) working on a version controlled local-first password manager for exactly this reason.
Keepassxc is local first and has password history. Check it out before building.
It didn't have a good sync story when I checked last.
Still does not. My approach is to keep the file in OneDrive. On windows / mac it's just a file, on android it's via custom onedrive protocol handler but also seamless.
Does it at least store the old password for a while in some archive, like most competitors do?
It goes in the “View History” section of a password entry, with an option in the 3-dot menu for “Clear History”. Not sure how long this is kept
Not at all.
Yes it does
At least on 26, the passwords app saves a history with previous passwords.
Yep. I get anxious when Safari starts to offer a new password for an existing account. Having access to previous passwords would be such great UX, but no, no such thing.
I'll believe this when pigs fly.
There's a 0% chance it will work. Most websites I've seen have one or all of:
* Force you to use email or SMS as a "second factor" to unlock changing password even if you know the old password
* A stupid idea of password complexity usually requiring one of a finite set of 5-8 "special characters" which is often only revealed after you've chosen a password that doesn't have them. Or in some cases even banning characters other than the ones they check for. There's a standard for this where you put a regex on the password field, which a good password manager will always use, but the kind of idiots who think limiting the entropy of passwords to increase security is the correct way to do things almost NEVER implement this.
* A maximum password length, even as short as 16 characters in many cases
* CAPTCHA etc.
Any effort spent on this would be better spent elsewhere, including even educating other companies on how passkeys should be used.
Some of your points are addressed by: https://github.com/apple/password-manager-resources
> Force you to use email or SMS as a "second factor" to unlock changing password even if you know the old password
Apple has detectors for codes sent via email or SMS, if your email account is one that is configured with the OS mail client.
> A stupid idea of password complexity usually requiring one of a finite set of 5-8 "special characters" which is often only revealed after you've chosen a password that doesn't have them. Or in some cases even banning characters other than the ones they check for. There's a standard for this where you put a regex on the password field, which a good password manager will always use, but the kind of idiots who think limiting the entropy of passwords to increase security is the correct way to do things almost NEVER implement this.
An AI agent can read the failure message and craft a new password
> A maximum password length, even as short as 16 characters in many cases
Same deal
> CAPTCHA etc.
While there's always the complex solution of scanning the image and trying to detect what is going on or slide the puzzle with enough of a curve to act like the motion of a human limb, there's also Private Access Tokens, supported by both Cloudflare and Google-provided captcha systems now IIRC. The OS uses an anonymous system to assert a single bit that there's proper browser chain-of-custody.
> Any effort spent on this would be better spent elsewhere, including even educating other companies on how passkeys should be used.
There are proposals as well to provide API to do upgrades from passwords to passkeys as well automatically. Nobody said the feature has to always use AI - but it may help the feature be robust enough for people to seek it out and try it.
Don’t forget those sites/apps that split the sign in process across five screens for bow good reason or those with mislabeled fields that password managers can’t pick up on.
I don’t think I’ve seen a single category of UX fail as hard and as often as auth screens do. It’s like at some point after 2015-2017 developers were struck with mass amnesia and forgot how to build decent login UIs.
Limiting the character set is done to reduce the frequency of "can't enter my password" support calls, not to increase security directly. Same with the maximum password length.
They may be limiting entropy to make it easier for users to remember their password. A user that can't log in is most likely one that will churn.
I don't think firms like the electric company or (payroll company) ADP are worried that I'll churn.
Also, the Venn diagram of "memorable" and "reasonably secure" really only intersects in the region of "Correct horse battery staple" phrases -- and the problematic sites I'm talking about nearly always limit length, which thwarts that type of password terribly. What is the purpose of maxlength on a password?? These shouldn't be stored in any form other than a hash, so unless long enough to pose a DoS threat during the hashing process, length is truly none of their business.
The entropy of a hashed password is limited by how many bits long the hash is.
1Password has been able to do this for five+ years. Frankly, it doesn't even really need agentic AI, although a talented team could probably make it perform better with agentic AI.
I don't really believe in Apple being that quality team.
I mean every api/app/website has a different way to do this. If there was a standardized api that everyone could conform to to allow this automation I would be all for it. I assume 1p does this by writing custom code/rules for dealing with the most popular sites out there and then erroring out for anything else.
AI could potentially help solve those unpopular site/app/whatever edgecase.
https://specification.website/spec/well-known/change-passwor...
> I don't really believe in Apple being that quality team.
Why?
Maybe observing Siri for the past 10 years?
They have no expertise in this area and their software quality as never been worse.
Tbh I assumed the Passwords team would be handling this and not the Siri team. Maybe I'm wrong.
its clear from their efforts thus far (image playground jesus fuck) that AI and even prompt engineering talent actively flocks away from them
Haven't you heard? Prompt engineering is dead. The cool kids are making Claude prompt itself. They're writing loops, not prompts. It's all about optimal tip-to-tip efficiency now.
“I don’t prompt Claude anymore. I have loops running that prompt Claude and figuring out what to do. My job is to write loops”.—Boris Cherny
heh I merely said prompt engineering because their efforts amounted to writing a prompt and sending it off to a model somebody else created to create some awful images
What's wrong with image playground? I haven't used it.
remember the first will smith eating spaghetti memes? thats the quality
I hope they don't feed the actual password into the model.
Don’t they say it runs on device? Then why not?
Judgment.
In Feb this year, when I analyzed KuzuDB's source code, I predicted Apple's reasoning to buy them was to introduce Siri with cross-app personal context. "..WWDC 2026 or 2027 introduces any “contextual intelligence” features in Siri that require cross-app relationship reasoning." https://medium.com/data-science-collective/i-analyzed-163k-l.... There is no confirmation on which tech is being used to achive that though
Microsoft cannot compete in browser race and has to adopt Chromium.
Apple cannot compete in AI and has to adopt Gemini
Google is a really amazing company.
Google is a great developer, but a really shitty consumer company.
Nest devices are garbage, it has been like a decade since their phones were competitive with other vendors, ChromeOS was barely hanging outside of education center and now it's a zombie walking since Apple released their cheap Mac Nano, Gemini is a joke as a product compared to Anthropic and ChatGPT. The only things worth something are Chrome, Gmail, Workspace and their Cloud.
Too bad Google can’t compete in consumer or professional AI either
What do you define as professional AI? If you're talking about enterprise AI I'd say they are well ahead of Microsoft and AWS.
In both cases, Google built a commodity that doesn't necessarily generate revenue for them. I don't think they're doing very well in this case.
Google's power is in controlling your interface. How you interact with the internet, and now, how you interact with AI, is dictated by them. There is massive power in being the interface between you and information.
Not directly, but they control the most popular portal to the open Internet with chromium to shape ad delivery, and (depending on how dramatic you want to be) the shape of reality presented to their users based on their AI models. Not directly profitable, but damn that's a scary level of power.
And it's amazing that they are doing very well overall without having to do too well to generate revenue.
[dead]
No, I don't want Apple to read all my emails, text messages and photos. I'm probably not going to buy Apple devices any more if it really works like this, by default sending all your data into their cloud.
If I want to use AI, I want to be able to select the exact messages / photos which I want to send to it. Otherwise I expect the device to keep the data protected. I don't need any of these features either; I can remember if someone sent me a cookie recipe.
You might want to so some research. It’s the only company with a level of privacy on this. Most of it is on-device. Private cloud compute does not store anything. I dunno man, you could criticize 360 aspects of this presentation but you pick privacy, really?
It can make on-device tool calls using your data (web search, database search, email, app APIs) which are not private. PCC doesn't protect against being dumb with your data.
E.g. "the user asks if their Bitcoin private key is unique, let's make a web search".
Oh damn I didn't realise it sent all this to the cloud. That seems kinda concerning.
For now at least, you can turn Apple intelligence off entirely in settings.
Apple is not reading anything.
Source: trust me bro
Nah, they actually have everything open for auditing and you can do that yourself. They heavily advertise that and had been for a while since they announced that split model 2 years ago
The best evidence is that they have never provided data to law enforcement. Ever. Because they can’t.
Private Cloud Compute is audited by independent third parties to verify it.
This looks pretty promising to me. It will likely replace the need to set up OpenClaw for average personal users. The work of getting email, messages, and all the personal data on the phone as context seamlessly is not as straightforward as one might think.
I'm curious how the pricing will work. Would it be free up to some limit and then some subscription pricing? I can't imagine it can be free unlimited usage given the price of serving these models.
Craig mentioned near the end of the keynote that compute intensive things (like image generation) will have rate limits that can be increased bundled with their iCloud + plans. I imagine any request that gets routed to their cloud compute will be subject to limits as well. He positioned it as a value-add to their existing subscription but I suppose that can change.
I doubt the average personal user knows what OpenClaw even is though Google is also producing competitive stuff.
This is why OpenAI thinks it needs to build its own physical devices. If Apple is only allowing its own AI to operate at the OS level, then that leaves OpenAI with no choice but to build their own.
Meta also realized this and attempted multiple times to build their own hardware but they've given up each time. They started as early as a partnering with HTC in 2011 to make a Facebook phone.
Quite frankly, I'm kind of excited to see what OpenAI can build. I think an AI-first phone could challenge iOS and Android. It's a new paradigm and if OpenAI gets it right, it'll be very hard for Apple and Google to pivot.
I personally think chat + code is the future of apps. For example, I find myself wanting to do many things inside ChatGPT instead of traditional app because I can tell it to do things that are simply impossible on a static app UI. For example, I have some data I want to send to an app but before I do, I want ChatGPT to clean the data in some way first. And then after the data is uploaded, I want ChatGPT to pull some data off the API and make charts that I want to see.
I imagine a world where very intelligent models run at 10k tokens/s, app building is extremely standardized, and it simply builds any app you want inside the OS. IE, if you want a dashboard of your health data, you ask it to build it almost instantly exactly how you want it. I'm already doing something similar today but it's slow and not easy to do for non-engineers.
> If Apple is only allowing its own AI to operate at the OS level
Incidentally, that’s what’s preventing Apple from rolling out their OS-privileged AI in the EU, as the EU mandates equal access for competing AI products. It will be interesting how this plays out.
Indeed according to Apple https://www.apple.com/newsroom/2026/06/due-to-dma-siri-ai-de... when allowing equal access for competing AI products
> Apple would have to give any virtual assistant direct access to users’ private data — and the ability to directly control other installed applications — as soon as Siri AI is made available in the EU
I don't personally agree with EU's mandates. I think it's ok if Apple only allows their own models to run on iOS at the OS level.
If OpenAI makes their own AI-phone, do they have to let Anthropic and Deepseek run their models on it too?
> do they have to let Anthropic and Deepseek run their models on it too?
provided it gets big enough, yes. the EU's position roughly is "if this hurts an entire market just to benefit you, and lots of people use / rely on it, then you gotta allow it"
Then the EU’s not gonna get anywhere the US and China will just sail along. The lawyers in the EU are essentially cutting their throats building hardware and operating systems cost real money and time which is one of the reasons Apple is using Google’s model for $1 billion a year. (when Apple finishes their models in house, Google probably will be dropped). Like dropping Broadcom, Intel, or Qualcomm in 2027-28).
It takes real time to drag along five ecosystems. That is the main reason it’s taking Apple longer than they’re so-called competition? Noticed that Google and Microsoft only do bits and pieces. Microsoft has no mobile and Google at its heart is an ad company the processor is six years behind.
that's the most common criticism. can't say i disagree.
though monopolistic practices are unequivocally bad and used to get struck down. one may argue this is just another instance of disallowing monopolistic behavior.
the DOJ used to have sharper fangs than what the EU is doing now
Why can't Anthropic or Deepseek take the big risk to develop their own phone? It doesn't seem right that they can simply use EU laws to hop on the ride for free without taking the same risks.
As a consumer it also doesn't seem right that Apple can just use all their private APIs that no other company is allowed to use to tell me what I can and can't use on my phone. If I want Anthropic to have the same level of access that Apple grants itself on _my_ phone, I should be able to do so.
Don’t buy it you have Android which is 3/4 of the world? What’s the problem? You also have Windows which is also 3/4 of the world.
As if Android and Windows don't also use the same tactics. Windows used to not let you uninstall Edge and Android used to just default to Chrome and Google Search, now they have to prompt you what you actually want to use. If the big companies are left unchecked, of course they'll use their size to bully other players to increase their own profit, but always at the cost of the customer.
Surely there must be a phone you can buy that lets you do that right?
If people want to buy open hardware, then just buy those. If they don't exist, make them yourself.
Do you rather want your refrigerator, washing machine and coffee maker send all the data it collects about you during your usage to some company you don't even know? How is it ridiculous to want to be able to make your devices do what you want to do, not what the big company behind it wants?
Those machines don’t connect to the internet but I want to run my own software on them.
1. A lot of those more recent machines do actually connect to the internet. 2. Since you want to run your own software on them, but you can't buy open hardware like that, did you build it yourself then as you suggested? "If they don't exist, make them yourself."
1. I want to run my own software even on the ones that don’t connect to the internet. /s
2. No, I don’t have the skills to. But if it’s so important to you, you can surely build it yourself.
Just make a new phone OS lol, problem solved.
If the demand for installing your own software is so great, a company would make it.
the EU feels like this falls under monopolistic practices, which it has deemed illegal. the buck stops with the politicians - there's no reason other than "the EU thinks it's bad for the economy and should thus not be allowed"
OpenAI is not a hardware company or an OS company and they hardly have the money to pay the real cost in the long-term, if Amazon, Meta, Google and the Linux crowd are having trouble, OpenAI is not gonna get the job done.
They have thier very own ai to help; isn’t that supposed to disrupt everything?
They would start on a green field though… which makes it pretty efficient initially
> Quite frankly, I'm kind of excited to see what OpenAI can build.
I’d go further and claim if controlling the hardware is a existential threat for these companies, their ability to deliver or not should be seen as an indictment of the entire “llm formulation of ai” era we’re currently in. If llms have the potential they claim, then empowering them to design a best of breed phone should be table stakes.
>Meta also realized this and attempted multiple times to build their own hardware but they've given up each time. They started as early as a partnering with HTC in 2011 to make a Facebook phone.
I was working in cellphone sales at the time and I can tell you no one wanted that phone back then even when Facebook was massive. An easy to hit facebook button was not a value add anyone was begging to exist.
Although with how many phones now have stock forced installs of Meta apps perhaps they won their con in the long game.
Similarly no one really wants a physical AI device, and attempts at such are pure techbro hubris on the companies part.
I think I want a physical AI device.
I want an AI assistant that I can use truly hands free. I keep my phone in my jacket when I'm riding my motorcycle. I want to be able to start, stop, adjust, and check details in route guidance. I want to be able to ask what the weather is like ahead on my route. I want to be able to ask it to start looking for a sensible place for me to stop for fuel and/or food without making me do a big detour.
Actually I would also quite like better driving directions, since I can't look at the directions on a screen.
None of these demands need an AI in a box.
They need software companies to give a damn and improve apple maps/gmaps/etc.
Or perhaps pull over and adjust your routing like most of us do, or maybe bolt a sacrificial phone with maps on it to a holder. If you don't want to risk your main phone.
> Or perhaps pull over and adjust your routing like most of us do
The entire point is I don't want to have to stop in order to receive new information.
I don't know how to make that clearer, sorry.
Todays pulling over to adjust planning is yesterdays pulling over to make a call, and yesteryears pulling over to throw a map on the hood.
Riding on a motorcycle is already dangerous enough with the average land tank driver on their phone. Talking to your AI assistant while riding at speed sounds like pending split focus disasters waiting to happen.
I think it’s time you hand in your licence.
For getting off the road to park somewhere and replan a route?
Are you a bot?
For apparently being unable to listen to or participate in a conversation while operating a motor vehicle.
Which is distracted driving.
Earbuds exist.
I use Shure se215 IEMs inside my helmet together with a Cardo Packtalk (which includes a microphone mounted inside my helmet).
The audio setup is not the problem.
The problem is when I say "Hey Siri, cancel route guidance," it will say something like "…You must unlock your iPhone to do that."
The entire point is that I can't take my gloves off, get out my phone, and unlock it, at speed, on a motorcycle.
Average personal users don't need OpenClaw at all...
The chatbots(ChatGPT, Claude et al) showed Apple exactly what can be done, the user base is already well primed. So this is a product definition done for them to execute. If done well they will be able to provide a much stronger integration into the day to day use cases than the chatbots, and can siphon off user time from them. This time around the end to end is easier with Apple Intelligence and more importantly llms doing the work Apple is floundering at. So I am hopeful, but I still see the os/app level integration as not enough in terms of functionality to make it a hit. The primary use case for llms is still conversations and search. Apple should be focusing on that aspect primarily and also add the os/app level integration as a bonus - as something only they can do. If they just do the latter, it will not be as much of a success. Let’s see how they execute.
EDIT: To provide meaningful chat functionality they have to either eat up the cost or charge a subscription for it. This will be first time they charge for Siri - a product that doesn’t garner any positive reviews. This gets even more interesting to watch
> This will be first time they charge for Siri - a product that doesn’t garner any positive reviews.
It seems like revisionist history to say that; lots of people were sold on iPhones years ago because of Siri. They have one of the few business cases for voice assistants, which are notoriously difficult to actually monetize, that actually makes any sense, since "selling iPhones" is meaningful and "selling a subscription" would be nice on top of that.
If they are using some mid model and are stingy with web search, I won’t use it more than a couple times.
> Aga sent you a message about Calanthea, a plant.
> Aga: have you heard of Calanthea? It’s a plant.
Really groundbreaking use of AI!
That's not what that is showing, it isn't reading a text you just got. It's you asking about the plant she told you about, it finding that text message, summarizing, and also proving with the text that it's accurate.
> Schiendelman sent you a message on hacker news clarifying what its showing. It's you asking about the plant she told you about, it finding that text message, summarizing, and also proving with the text that it's accurate.
wow! we are in the future!
I chuckled, thank you. :)
I've not seen anyone discuss yet what kind of impact these AI features are likely to have on battery life. If they are running small on-device models, this will have a measurable impact on battery life with regular usage.
Before they add AI they better fix the frigging search function in settings, it is horrible, you need to know their exact words, and Apple has a funny naming sense. Hierarchies nested so deep you never find anything. I come to use Claude or ChatGPT to tell me the right incantations to find a setting.
I don't even know that it's knowing the exact words, it just seems like sometimes the search decides it's just gonna give you no results because the vibes were off.
I'm almost sure that sometimes searching the same thing will give you the result and sometimes it won't.
I wonder how much of Siri AI is Apple-developed and how much of it is Google-developed as a result of Gemini. The a) search demos and b) image generation demos seem unlikely to have been done by Apple alone, the demos being closer to Google Search and Nano Banana respectively.
It looks almost entirely like gemini. The images they showed are obviously nano banana, and the text responses are almost obviously Gemini (I say as a somewhat frequent Gemini user).
I'm sure they customized some of it, but this looks basically like Gemini integrated with iCloud instead of Google Workspace.
I'm curious what the obvious tells were for you. I never use any of these tools so I do wonder what sets them apart for those in frequent contact.
The images were the biggest tell, generating using a reference photo of a person, at least Gemini and ChatGPT have two distinct styles. ChatGPT is a little less uncanny valley than Gemini which tries to be too realistic looking, in a bad way because it tries to preserve the original person in the photo, but still can't seem to help altering facial features.
The text responses had Gemini's verbosity. Asking ChatGPT to show me iconic dishes from both Brazil and Morocco (Apple's example), is much cleaner, less verbose. Quick list of dishes and links to the recipe. Gemini just spews a wall of text and bullet points and goes on and on with fluff. Tons of "What this dish is" "Why it works" Same with its frequent use of tables, which I see less of with ChatGPT.
Each Siri demo they did in the keynote had that hallmark verbosity I typically get with Gemini without prompting it to not do that.
Yeah I noticed the same things! I've been using Gemini on my Pixel a lot like this and this feels like Siri skin for Gemini
There was not a single thing they launched that I have not seen Gemini already showcase capability or existing feature wise
The only thing you saw was phoning home, that’s what’s gonna be interesting when Apple releases their version. Is it phoning home 100% of the time or can you turn off the Internet and have it perform in the same way, there will be plenty of YouTubers that will give it the test a test I might add that they haven’t done up until this point for anything that Google has put out?
Is Siri any more or less than “just” an agentic harness such as OpenClaw? How much of what that harness does is up to the LLM or the harness itself?
In my mind the Gemini LLM defines the bounds of capability and capacity, but any actual functionality or usefulness (or lack of) comes from Apple’s Siri harness.
I was wondering the same. I have to imagine it's mostly Gemini, unless Apple has a big, secret, SotA foundation model no one has heard of? But if it is Gemini, how does that work with their Private Cloud thing? Are they able to load the Gemini weights into it?
IIRC Apple cut a deal to have their own version of Gemini that is hosted just for them.
It’s almost all Gemini and the “Apple local models” part seems to just be image embeddings/descriptions powering new spotlight and the like which is also likely someone else’s model.
To this day, Siri does not understand "Subtract 5 minutes off my timer" when having an active timer
In English!? Someone please Apple that LLMs can deal with multiple languages at once without the old “go to Settings to configure your language”
I wonder if the language support is the voice part of the assistant? It took a while (years) for Siri to speak my native language back then.
Not with the EU, support isn’t coming the lawyers in the EU won’t get there before the engineers in China.
They should have changed the name as per branding. I hear Siri, I subconsciously associate it with really bad software.
Hey if they could turn Apple Maps around, why not Siri?
I hear Apple and I associate with it with laughable AI integration, an unserious UI, and buggy software.
Apple’s entire software stack has a branding problem.
This sounds like a take that you'd probably only find on HN though. It seems dubious that regular Apple users see any issues with the software stack or with Siri, certainly not to the point that it'd be a branding problem for Apple.
Next time you're speaking to a close non-techie friend or family member feel free to ask them how often/whether they ever use Siri, and then ask them why not, and you'll realise that no, this isn't some HN-specific peculiarity.
I don't have to wait for a next time: my mother uses Siri literally all the time, it drives me crazy lol
> and you'll realise that no, this isn't some HN-specific peculiarity.
I think it is tbh
Quite possibly one of the most lackluster WWDC's I can remember. And I've been a fan of Apple/Mac since the late 80s.
Hmph.
That said, I'm THRILLED they claimed to "fix" the border radius snafu of Tahoe. Go ahead and push that now with the next Security fix. We won't mind at all.
I placed eggs into boiling water and because my hands were wet I used voice and said "hey Siri, start a timer". It replied "you'll have to unlock your iPhone first", so I hovered my face over the bench the phone was sitting on, then the screen rattled and prompted for the numeric passcode. Ugh..
Using Siri essentially required me to use my hands anyway, so what's the point of voice?
I'd very seriously consider moving away from iPhone to a device that treats voice AI as a first class citizen (presently I mapped the 'double back tap' to open grok voice chat, and triple back tap to end it, which is a wonderful improvement over not having these, as you can do that pretty easily, even while driving etc).
Don't do back taps require hands?
They do (one hand), but it doesn't require anything precise (like finding the app and clicking on it, which you can't really do while driving). If Siri could open an AI voice chat, that would awesome. But I'm not counting on that within this decade.
The only way Siri will meaningfully save my time would be if it could doomscroll for me. All of the things Siri claims to automate here would save me maybe a couple of minutes total over a week, wouldn't make the tiniest dent in my screen time.
> if it could doomscroll for me
I don't understand. Are you so little in control of your own self that you now need an embedded AI to do something for you that was already a complete waste of time and energy? Just delete the apps.
The point I was trying to make was that consumer AI on mobile devices won't save people time in any significant way because people don't spend a lot of time on their phone doing things that are automatable. Everything mentioned here is a tiny portion of people's phone use compared to things like doomscrolling, video calls etc. So there is no real value to any of these tools.
What is not demoed are basic AI capabilities I need today to effeciently use my iPhone. Examples: - Siri, send the latest 20VC podcast to my watch. I'm going jogging. - Siri, do not start Apple Music. I do not use that app. Start Podcast. - Spell check and Auto-Correction on iMessage should support French and English. I switch from the 2 all the time. - Etc...
I'm sure you guys have tons of others to list.
I read through the entire DMA rant that apple has here: https://www.apple.com/newsroom/2026/06/due-to-dma-siri-ai-de...
This feels like it could be solved with a list of permissions that the user has to turn on when using 3rd party AI.
Apple already:
I feel like this is nothing more than Apple being angry that they have to allow people to actually choose what AI they want on their phone. This is particulary interesting if anthropic and openai decided they want to add siri ai override to their apps allowing them to take advantage of the apple ecosystem without signing some kind of deal like they had to with Google. I assume behind closed doors Google had to make some sacrifices for them to be the model powering siri.> I feel like this is nothing more than Apple being angry that they have to allow people to actually choose what AI they want on their phone.
It's really just Apple being angry about the EU's DMA endangering their golden goose (App Store revenue) and using any meaningful new functionality as a bargaining chip.
They've done staggered geo launches for other features in the past many times, both before and after the DMA was passed, and in this case there's even another great reason to not want to globally launch all at once (AI inference server capacity). If they can at the same time market it as part of their ridiculous turf war against the European Commission, I guess they just have to take the opportunity.
The thing is, Europeans are mostly annoyed with Apple over this, not the EC.
It just reads like arrogant foreigners throwing a tantrum over our laws.
I struggle to believe there isn’t a significant fraction of EU citizens who are frustrated with the EU’s laws.
At the very least to the extent that the whole setup limits national sovereignty.
Your comment comes across as though you expect us to believe EU citizens are a homogenous whole, who happen to align with your perspective on this matter.
As someone who’s now switched over completely to USB-C and has saved money and reduced cable complexity incredibly, even if I ever get mad at the EU it will never be over Apple, which has time and again demonstrated itself to be an incredibly bad faith actor.
They complain about things their competitors are able to implement with no problems at all, and then once their tantrums are not sustainable anymore, they’re apparently able to solve all the problems as well.
For a company that sells itself as a design + engineering firm, they seem to have very little confidence in their ability to design or engineer their ways around constraints pretty much everyone else is able to.
On the contrary, it reads to me like they've simply been around, and this is the general impression they gathered. Which may still not even be true, but it makes a whole lot more sense context wise, and is pretty darn different to the conveniently malicious motivation you're proposing. And with this, now your own "royal we" is similarly rendered deceptive.
> At the very least to the extent that the whole setup limits national sovereignty.
That's how anything grouplike works indeed.
> I struggle to believe there isn’t a significant fraction of EU citizens who are frustrated with the EU’s laws.
Sounds like something that'd have polling data coverage?
[flagged]
We've banned this account for breaking the site guidelines and ignoring our request to stop.
If you don't want to be banned, you're welcome to email hn@ycombinator.com and give us reason to believe that you'll follow the rules in the future. They're here: https://news.ycombinator.com/newsguidelines.html.
[flagged]
Please don't respond to a bad comment by breaking the site guidelines yourself. That only makes things worse.
https://news.ycombinator.com/newsguidelines.html
[flagged]
Supranational regulations limit national sovereignty, news at 11.
Across the gamut of regulations the EU has, it's not really the ones that apply to Apple that draw much ire.
The alternative is more US meddling.
I would call those people uninformed. It's completely reasonable for Apple to be hesitant to roll out any new integration features within their own ecosystem in the EU.
iPhone mirroring for example. Seems like practically 100% chance that if they put that out in the EU they'd be facing lawsuits for not making it work with every Android and Kindle and digital pregnancy test on the planet. And making it an open API right out of the gate is a much bigger undertaking than just making it work with your own devices through a proprietary API that you're free to break at any point and just update your devices to work accordingly.
Ignoring your silly hyperbole—
Yes, it is harder to support APIs compared to… only having a completely closed system.
It would take a trillion dollar company to have the kind of resources that kind of thing would take.
The fact Apple thinks it should be exempt from rules that try to impose some fairness and choice, simply because “it’s hard” is the kind of thing that makes me feel they should be forced to be broken up in some fashion that limits the conflict of interest. They can be a petty platform owner who extracts rents from every developer OR they can offer things like apps and services on top of a platform. They are proving themselves too evil to be both.
> [...] for not making it work with every Android and Kindle and digital pregnancy test [...]
If only there were a way to design a system in a way that didn't even require you to know who's on the other side of an interface. Some kind of... protocol specification maybe.
Imagine the things we could do, like making a phone call to a different brand phone than yours, or a web server running a different OS than the client, maybe even one that the other side has never heard of!
It's hard to take your argument seriously when you're being hyperbolic.
> Given the serious risks to users, Apple designed a solution called Trusted System Agent — an intermediary that would allow virtual assistants to safely access the same features and capabilities as Siri AI for devices in the EU. Apple also shared a plan to launch Siri AI in the EU while gradually rolling out this new solution over an 18-month period. The European Commission said no. In fact, the European Commission did not agree to any of Apple’s proposals.
I'm extrapolating (there is less detail in that press release than I expected from your comment), but this sounds to be like it would be the thing that enables such a "list of permissions". I would be curious to know exactly what this agent entailed and why the EU did not approve it.
"Trusted System Agent" imo sounds like an apple approved agent which would only be available to companies that accept apples (likely unreasonable) demands and would completely lock smaller companies out of the ecosystem.
My limited understanding is that it would be a local model that exists only to determine a limited set of local information necessary to answer the user's request. This request and information would then be shared with the third party. Third parties would otherwise not have access into the local semantic model based on user personal data.
Does apple's own siri need to pass requests through their gatekeeper AI? I bet it doesn't. Personally, I'm generally happy with any answer apple comes up with so long as they're bound by the same set of restrictions as 3rd party companies. I feel like that's the only way to make sure apple won't "accidentally" hobble their competitors. (Like they did with their ridiculous 50c per app install fee for 3rd party app stores).
I want Claude and OpenAI to be able to compete with Siri on an equal footing. Just like Apple maps has to compete with google maps. Competition pushes companies to make better products.
Hm, I didn’t even consider that it could be an “agent” in the AI sense. I assumed this meant a service that runs on the device and interposes on requests to access privileged resources and enforces permissions checks on them. That is, the classical sense of the word agent in computing. Perhaps you’re right; in any case I don’t think there is really enough detail here to go off of.
It’s kind of funny that the EU’s regulation here would force Apple to allow options that are worse for user privacy. Apple is the least incentivized to farm data from its users; in fact, that’s a huge selling point. They mentioned it over and over and over in the WWDC keynote today.
In my opinion, Apple is doing the right thing for users. It’s not like they have a huge revenue stream here. Yes, there will be some features or usage that require iCloud plus or whatever to cover incremental cost, but I genuinely believe that they don’t want services creeping in that break their trust with users or their privacy-first reputation.
Apple’s decision (users will have a less powerful product because we’re not vacuuming up their data and using it for profit) is exactly the kind of thing the EU should want. No country has appropriate data privacy guidelines for AI (yet) so opening up choice can’t provide alternatives.
(To be clear, I’d be fine with Anthropic here, but am fine with this state. Maybe because I’m so used to Siri sucking that I’ve given up hope.)
> Apple is doing the right thing for users.
The right thing for users would be to allow user choice, and for Apple to compete fairly.
Apple allowing third party access doesn't automatically mean user data gets hoovered up by OpenAI, Anthropic, etc. It just means users now get the choice, if they want to make that choice. Users could stay with Siri/Apple if they care about what Apple is offering, or choose to accept the risks and terms of service with other third parties.
The EU isn't saying "you must preinstall every competitors offering" its "you must offer the ability for others to hook into the same APIs to be able to offer their own assistant on par with the first party option."
The user still remains in control by virtue of their own choice.
I never understood how any regulatory body is going to decide which APIs in iOS must be made available to third-parties to hook into. So what if I'm a third-party maker of TCP/IP stack and I want Apple to offer me the ability to sell my custom TCP/IP stack to my iOS customers as a replacement for the stock TCP/IP stack that ships with iOS. Clearly no regulatory body has cared about that because it's too niche of a space?
So some government official will scour the entire API surface of iOS and decide which ones Apple needs to expose to third-parties? They have already decided App Store and Payments APIs need to be made available. Now it looks like they also expect off-device foundation models need to be made available to third-parties.
What about making Apple Watch specific APIs in iOS be made available to all third-party watch makers so any one can bring any smartwatch and use it just as effectively as the Apple Watch with an iPhone? What about all the AirPods specific APIs that lets Apple offer a better experience with AirPods than a generic bluetooth earbuds? What about Apple Pencil? And so on... If you go down this path, the list is endless.
> What about making Apple Watch specific APIs in iOS be made available to all third-party watch makers so any one can bring any smartwatch and use it just as effectively as the Apple Watch with an iPhone? What about all the AirPods specific APIs that lets Apple offer a better experience with AirPods than a generic bluetooth earbuds? What about Apple Pencil? And so on...
Don't threaten me with a good time? All of those seem like great policies. The fact that I cannot use an apple watch with an android phone is ridiculous, and vice versa as well.
Should you be able to use a Samsung SoC in an Apple phone?
At some point this is just a debate about vertical integration. Apple can deliver better experiences with it, but of course it limits user choice.
Many people want fully modular, open systems, which is lowest common denominator.
I can see both sides of the argument, but I am so skeptical of regulators deciding what can be integrated or not. If modularity is better for consumers, why don’t they prefer modular systems?
At the very least I think there should be a very clear tradeoff; right now the EU seems to think they can regulate their way to all of the benefits of vertical integration while outlawing vertical integration. I don’t see how anyone could look at that with a straight face.
> Should you be able to use a Samsung SoC in an Apple phone?
How did we go in less than two comments from providing access to APIs that are already present, implemented and actively used by Apple (who in their holy wisdom deem us mortals not worthy to access these the way we choose) to a completely different hypothetical of requiring actively building support for another companies hardware?
Such slippery slopes really aren't helpful, nor in any way comparable to what the DMA actually intends or states.
> why don’t they prefer modular systems?
Because there aren’t any to choose from?
“Smartphone” has become a mandatory thing everyone is required to use to function in society without major friction.
Businesses hate supporting a ton of distinct platforms, as proved by the developer marketplace killing Windows Mobile through refusing to ship apps for it.
This suffocates any third entrants just like the FPTP voting system suffocates third political parties.
So what modular OS are people supposed to choose?
> Should you be able to use a Samsung SoC in an Apple phone?
That's irrelevant to the discussion. SoC is not a digital platform in any way under DMA. It's not a platform at all.
Yes I'd like some of these too but at the same time I get an uneasy feeling when I think that some potential idiot in a regulatory body in every country is now going to decide which API surface needs to be made available to third parties. If they take it too far, they could end up making nonsensical choices and kill innovation.
> the list is endless.
Good. They are making an operating system. User choice and competition matter. I know Apple would prefer to allocate more resources to Liquid Glass animations and burying more UI elements inside “…” menus, but I personally think I don’t need any more innovation above the OS level from Apple. Especially because 80% of their changes to the application layer in 10 years have just made their platforms worse.
Let them ship a stable platform that allows applications to do tons of useful things, even when you don’t accept a mega-package of apps and services all from the first-party vendor that locks you in.
If Apple built houses, you would have to jump through hoops every time you use a microwave or lamp you didn’t buy from Apple.
> What about making Apple Watch specific APIs in iOS be made available to all third-party watch makers so any one can bring any smartwatch and use it just as effectively as the Apple Watch with an iPhone? What about all the AirPods specific APIs that lets Apple offer a better experience with AirPods than a generic bluetooth earbuds?
That's...literally the point of these regulations. Sounds great to me.
Creating competition where it would not otherwise exist is the essential nature of the EU. Originally it was mostly about forcing protectionist member states to accept competition from other member states. But they extended the approach to breaking perceived natural monopolies a long time ago.
The exact rules ultimately don't matter, because the EU is after outcomes. If the current rules don't lead to the desired outcomes, they will keep changing the rules, until they get what they wanted. (Or until their goals change.)
Destroying competition by removing the consumer choice for vertical integration in service of strong security, privacy, reliability, etc.... is mistaken.
It's competing at the wrong level.
The iPhone is a toaster. Nobody's up in arms about whether the toaster takes other manufacturer's crumb tray. It's a television, and nobody's demanding QLED and OLED be swappable. It's a console. Xbox doesn't play PS5 games. It's fine.
There's no real line between hardware / firmware / software / malware ... For what Apple offers consumers, every layer of whateverware should be trusted.
Drawing imaginary lines based on the embodiment or substrates for logic gates is mistaken.
There are lots of phones. Lot's of different philosophies. Stop taking away consumer right to pick a philosophy and design for an end to end experience. It's fine.
Nothing about allowing others equal access to the OS means that someone can’t still choose Apple’s first party services and products.
It’s not an either/or thing, it’s about preventing so called gatekeepers from anticompetitive behavior via favoring their own accessories and services while simultaneously preventing any others from possibly competing.
There’s no valid reason at all a third party smartwatch shouldn’t be able to integrate to the same level as an Apple Watch. No reason third party Bluetooth earbuds shouldn’t be able use ADWL for automatic device switching, etc.
Want to still use only Apple? Great, nothing says you can’t. But at least it would be user choice and there would be actually competition which would lead to better products for all.
Can’t believe I lived to see the day that people on HN start defending vendor lock in and closed platforms as a good thing. Have all the hackers retired?
We already have user choice - it's called buying an iPhone or not. Nobody is forced to buy an iPhone, or a smartphone at all.
The EU is just a disaster with these laws. They can't innovate so they do this.
> Want to still use only Apple? Great, nothing says you can’t. But at least it would be user choice...
It's already user choice. The problem is too many users like the lineup. And too many who aren't going to use it, don't.
I think if you actually invested time into researching the DMA you will be able to understand why they are making certain decisions.
Oh, me, me! I spent a few years being responsible for a significant bit of DMA review and CYA and responses to regulators.
I’ve read all of it, multiple times, and been grilled by EU regulators (vicariously, via corporate lawyers).
It still boils down to general guidelines that it’s impossible to know if you’re violating before the fact, and they will not even approve/reject proposals in advance. It’s basically “go read the act yourself, and ship what you think is compliant, and you’ll know whether we interpret the words the same way by whether or not we fine you.”
Good times.
> It still boils down to general guidelines that it’s impossible to know if you’re violating before the fact, and they will not even approve/reject proposals in advance. It’s basically “go read the act yourself, and ship what you think is compliant, and you’ll know whether we interpret the words the same way by whether or not we fine you.”
Companies want to know exactly where the line is so they can figure out how to comply with the letter of the law while doing as much as possible to get around the spirit of the law. This has been demonstrated over and over again. It isn't the job of the regulator to help companies with this process.
So you’d be cool with speed limit signs that said “hey, don’t go too fast” and no specific limits? And the cops decide who to pull over on vibes, reputation, mood?
I’m more of a rule of law person myself. If there’s a law that must not be broken, and breaking it results in penalties, it seems insane to me to not specify it in advance.
Sure, big tech is largely evil. Arrest ‘em, find them, IDGAF.
But pretending that DMA and related regulations provide enough information to ensure compliance is willfully ignorant. The regulations are designed to allow selective enforcemen.
https://en.wikipedia.org/wiki/Advisory_speed_limit#Germany
>speed limit signs that said “hey, don’t go too fast”
Yeah we already have that. We have the words 'SLOW' on the road that ask you to slow down from the current limit for the hazard ahead, but pulling you up on this is officers discretion.
Bad example, as there is a posted speed limit above which you are positively breaking the law. Discretion may lower the limit.
The DMA doesn’t have the objective measures. It’s all discretion, all subjective, all post hoc.
Which, cool, some people like the idea that police target those people and need flexibility to make life harder for undesirables in ways they would never do to high status people. I don’t personally like that, and I don’t think tech regulation should work that way.
Replacement TCP/IP stack sounds like a VPN—which iOS allows
VPN is not a replacement TCP/IP stack. I literally meant the TCP/IP stack in the XNU kernel. It might be an esoteric example but it's not that far off. DMA already forced Apple to open up browser engine layer so third-parties can now bring in their own browser engines in the EU and are not restricted to using just WebKit.
True. Will add, device must be supervised to use VPN always-on which is possibly sensible albeit annoying (would have to reinstall iOS and set up as new I believe).
> Apple allowing third party access doesn't automatically mean user data gets hoovered up by OpenAI, Anthropic, etc. It just means users now get the choice, if they want to make that choice.
Apple is also restricted in the sort of consent prompts they give the user. That could matter when a non-technical users is prompted by a third party app to effectively allow unfettered access to all user personal data on the device.
Sometimes when you look at the functional requirements for a feature it turns out to be a bad idea. In the EU, functional requirements can come after-the-fact from regulator interpretation of the DMA. Until Apple determines what those requirements actually are going to be, releasing a potentially harmful feature is irresponsible.
And why is that a good thing? The average user can't even spell Anthropic. Why do you think they can safely pick a third-party model provider that could harvest the hell out of their conversations? The control of ecosystem is part of the privacy and security. My mom's Android phone has like 100 apps that she had no idea how they were downloaded. For real user choices, the vast majority of users just want a phone that they can trust and don't have to be a techie to avoid being exploited. They can choose to buy a phone that can be built from legos, OR they can choose to buy a phone from someone they trust to get the privacy and security taken care of for them. This is the real user choice.
Forcing choices is not "competition". We already have competition. People can buy any phone they want. They can use whatever AI they want.
There's nothing "fair" about this at all. It's a group of luddites in the EU who dislike how successful American companies are.
Again, the EU is stifling innovation with these backwards-looking rules. No wonder they have no innovative companies.
There is a vast asymmetry in knowledge and capability to make those choices for most users. Most users will press agree and consent to things they don’t understand to get to the next screen, and while that would be terrible for the individuals I think it’s also important to look further at the ramifications because probably Apple would be blamed.
They can't under GDPR. The DMA is for market access - there are other laws for privacy. Those require use commensurate with what is needed for the service, so anyone who e.g. scraped all of a user's local info and stolen it would be breaking EU privacy laws themselves.
This is not complicated. Even in the US, every other industry is regulated to your benefit, you're just used to it and haven't realised. Digital technology obviously needs to be too. And yes, you have to do it properly.
Apple ad revenue is ~10% of rev, with Google deal, and growing. New management is going to turn it less privacy focused company, because Apple needs to pursue growth.
Yep and more like 25%+ of profits (given the google revenue, and most ad revenue, is close to 100% margin).
Apple may have a true argument that their version of “AI” is currently the least privacy-problematic, but it’s not a compelling one. First of all, it sets an awkward precedent. This special status is mostly based on vibes, and on how we know the other guys’ business models will push them towards more data collection. But doesn’t make sense to say we allow anticompetitive behavior depending on how nice to how “nice” the aspiring monopolist is to their users.
Google can easily argue that if Apple gets to rule over a walled garden, zero-API ecosystem where no one else can compete, then it’s right that they can too, regardless of how privacy-respecting they are or aren’t.
if for a second you believe that what apple says the regulators told them is the same thing as what the regulators told them, i have a cow farm under the titanic to sell you
This comment casts aspersions while making zero specific claims of wrongdoing. If you have something specific to say that goes beyond the vibes of "everything and everyone is corrupt and evil," that would at least be worth hearing.
Apple continually lies to the EU and the public about their DMA compliance efforts. Is that specific enough?
oh, it is worth hearing. said another way: "show me"
Your premise is incorrect; if apple truly wants to do the 'right thing for its users', it would allow choice. The fact that the current crop of likely alternate choices include quite a few companies and offerings that seem far more user hostile than apple's offering doesn't change that fact (it merely raises separate concerns that there need to be more laws such as the EU's DMA, not fewer).
However, even if your premise is correct, it does not matter.
In the end, trying to manage such products (require massive investment, have network effects, offer significant gatekeeping and rentseeking opportunities) is extremely problematic.
On one hand, the market cannot do it properly: There are tons of externalities, and, like e.g. building out rail, the absolutely gigantic barriers to entering the market means the existing players merge into a monopoly or oligopoly.
On the other, the product is too complex and too dependent on continuous evolution to officially turn it into a state-controlled / state-run monopoly (the solution many countries have deployed to solve e.g. how rail, or medical insurance, or road networks, end up in a terrible state if left up to the market).
So what is one to do?
The current crop of mostly US led large companies seem to have gone with a 'just trust me, bro!' argument, with some 'AI is so important you cannot put up any roadblocks at all!' sprinkled in.
And yet these companies time and again prove that they can't be trusted. Which is obvious and logical: Companies must conform to the law, but are otherwise amoral. Or rather, their 'moral' compass has nothing to do with human moral compasses: They must earn money for their shareholders, in whatever legal way they can find that is most efficient, paying as much attention to future company growth and health as its shareholders desire. That isn't just 'what they are incentivized to do' - that is what they are legally *required* to do.
And yet you've gone with a motif of 'but apple is the one company that is doing it right so lets just trust them.. bro'.
There *is* a solution:
Use the fact that the state has powers of persuasion that companies simply do not have. The threat of law, and the monopoly on violence.
Essentially, a state can simply tell a company: The populace have spoken and they value X (say, privacy). They value it a lot. You will deliver. At low cost. This is not a request, it is a demand. If you don't want to or can't, then we shall write laws to regulate you and then *everybody loses*.
Conceptually this works, in a weird game of chicken / madman theory: If the corporation in question believes that society will regulate them into oblivion unless they comply with society's demands even if this means society incurs a great cost, then the corporation *will comply*.
This has happened before. There is no actual law in the US that a movie gets a rating, and the movie industry pays for and manages the ratings of its movies entirely as an internal affair. And yet, in general, movie ratings are stellarly well run compared to what a government run institution would have done.
The reason *is* that threat. The movie industry decided to police itself because it was quite clear that if they did not, the government would have, at great cost to the movie making industry (and at significant cost to society as well, in the form primarily of much worse films).
For some reason that isn't entirely clear to me, CEOs of large corporations that deem themselves 'IT companies' do not understand this part. They will fight tooth and nail to fight every law, and especially in the US, perhaps due to extremely dire and long-term distrust by its populace in its own government, many of its citizens incorrectly side with its corporations on this idea, even though time and again corporations prove that they have no allegiance other than to the almighty dollar (which, to be clear, is not a complaint. That is how society has set them up. My only complaint is that e.g. you seem to have forgotten that this is how it works).
Hence, given that the system works on, in essence, fear / coercion, the only right answer is to do an attitude adjustment, find a massive club, and beat a whole bunch of IT companies into absolute pulp until the remaining CEOs understand.
And before you make a note about the brash, medieval nature of that comment - it is already clear that these CEOs who think they are God's Greatest Gift To This Planet, are already meekly running, tail between their legs, to kiss the pinky ring of a personalist wannabe emperor president. They are _clearly_ motivated by such fear and _clearly_ cannot be trusted to rise to the occasion and be a new form of benevolent leadership for the citizenry.
I wish they were. It'd be so much easier.
Siri AI is not just a chatbot. There are deep hooks throughout the entire OS and across third party applications. Siri AI has been given wide access to the user's semantic index which will encompass just about everything on the device. Also remember that it isn't just the Apple user's privacy at stake. Any substitute will also have access to any interaction with other people. Apple is at least claiming to not keep any data people use with Siri AI, will Google, Anthropic, Open AI, etc. pinky swear they won't build a profile on me because a friend of mine chose their AI over Apple's?
I will wait and see what people find out about it before passing judgement. It's quite possible that it isn't possible to have an API to use other companies' AI instead of Siri AI. Are there any equivalent API hooks on Android?
> This feels like it could be solved with a list of permissions that the user has to turn on when using 3rd party AI.
Nah, that just shifts the goal posts. If they did that, developers would be whining about "scare screens", as we have already seen when Apple put app installs behind a permission prompt.
They're already up in arms about the requirement from Apple (and Google) to know who is behind the apps that slurp up all your data.
The DMA maximalists won't be happy until Apple releases an anonymous service to automate setting up a Kafka topic to send each iOS user's PII to whoever wants to receive it.
You can downloads millions of things for your computer without kyc protocols. Why are phones in a special class? Your data is being slurped by the people who sold you the phone and you are worried about the small fish.
> Why are phones in a special class?
Because we applied the learnings from computers and phones have ended up way more personal?
> This feels like it could be solved with a list of permissions that the user has to turn on when using 3rd party AI.
The device won't be able to ask for significantly more permissions than Apple asks for their own model for regulatory reasons, nor will it be able to convey the seriousness of granting the permission (e.g. immediately give unrestricted access to the vast majority of personal information/documents stored on the device).
But Apple also architected their system to justify not having constant permission prompts for access to sensitive data. And for regulatory reasons they also can't mandate that competing models have the same architecture.
The regulators and Apple (along with hopefully other AI companies) will need to work together to determine longer-term stable path forward.
Apple could have the same kind of permission dialogues with their own models (and they actually should). Each and every (first-time) use of a feature should:
Putting this in practice: In this case: 1 is access to data, you might want to give broad access to some applications and input data2 is permission to act, but you might want to deny access to some parts such as sending email and scope to summarization
> "requires developers to submit ID to publish an app on the appstore (at least I had to after ~1000 downloads to be able to publish an update)"
What is the purpose of that?
So there's someone to sue if the app misbehaves.
In my opinion, this is late by almost two years. This should have been the v1 when they first presented Apple Intelligence.
Coming this fall...
In "English" later this year...
We've heard that before, haven't we, Apple? I feel the right way to fix the trust issues would be to announce this when it's actually done. Like, here's Siri AI, and you can download and use it, right now.
You can. Install the beta
EU. iPhone 13 Pro. Not complaining, but I can't really.
So Siri is basically now a Gemini agent?
Our family uses Siri with a HomePod a lot, and it's already much better than it was a couple of years ago where it could basically set timers, tell you the weather. Now it answers questions ("when did the Knicks last win an NBA championship") with decent answers, instead of "I'll send the web results to your phone". But it's still far behind voice-chatting with Claude in the Claude app, so very much looking forward to this upgrade.
I will say though that proper voice transcription in Claude -- or any of these agents -- sucks. If it can't understand the question properly, then it can't provide the right answer. It works okay for me, but not for my kids, not when speaking quickly or in incomplete sentences (as people tend to do), etc.
This whole "coming this fall", "later this year", it's annoying. I miss the days when Steve Jobs used to say "and it's available right now, you can demo it in the hall outside, we're going ot make a billion dollars by tonight."
So... 2001? Because by 2002 (MacOS X 10.2) Steve was already "previewing" the next MacOS at WWDC for a release in the fall.
To be fair this is the conference for developers - not the general public
> Available on iPhone 17 Pro, iPhone 17 Pro Max, iPhone Air, iPad models with M4 and later and at least 12GB of unified memory, and Mac models with M3 and later and at least 12GB of unified memory.
It’s really disappointing to see the on-device models being limited to so few devices. And this was after the iPhone 16 and 16 Pro were marketed so heavily with supporting their now failed effort at AI.
Apple has pulled a Tesla here. FSD on HW3 cars is stuck on old software with no upgrade path as of now. Tesla is potentially justifying it by calling it "FSD (Supervised)" so they don't have to do an expensive retrofit to them even though they sold these cars originally with the promise of fully autonomous driving.
All the iPhone 16/Pro owners have been waiting for Apple Intelligence features announced from that WWDC 2 years ago. They didn't get delivered and now won't ever be delivered with on-device intelligence due to the 8GB RAM limitation.
I’ve got a 2023 Mac Studio M2, and was dismayed by the M3 & later. So I’ve been trying to track down more details. That specific device list is only for:
> Apple’s most powerful on-device model and the features it enables, like expressive voices and more advanced dictation, […]
On other devices, I think there’s still on device support (just not with the “most powerful model”), for these devices:
> Apple Intelligence and Siri AI in iOS 27, iPadOS 27, macOS 27, watchOS 27, and visionOS 27 are available on iPhone 16 models or later, iPhone 15 Pro, iPhone 15 Pro Max, iPad mini (A17 Pro), MacBook Neo (A18 Pro), iPad models with M1 or later, Mac with M1 or later, Apple Vision Pro, Apple Watch Series 9 or later, Apple Watch Ultra 2 or later, and Apple Watch SE 3 when paired with an Apple Intelligence-enabled iPhone nearby.
This is from the footnotes on https://www.apple.com/newsroom/2026/06/apple-introduces-siri...
I do wish they’d been more clear about what the “advanced features” are :(
That's specifically the secondary, more-powerful model. It was only mentioned in passing in the keynote, but on this page anyway, it seems to be just the improved dictation in Siri and ability to customize pacing, etc.
I'm glad I've been sitting on my iPhone 15 Pro Max... I'll upgrade someday, if/when I need to and the software updates are compelling, but I'll see how things run on my M5 Macbook. But the 15 Pro Max isn't subpar in any other way.
The 12gb number is weird, but also telling.
iPhones have 12gb, current Neo has 8gb, the next gen Neo is speculated to have 12gb (as it'll be based on a later iPhone chip).
> It’s really disappointing to see the on-device models being limited to so few devices.
At first I thought it was the usual planned obsolescence. Then I realized it may be a true technical limitation. I suspect an embedding model is required to run on device in order to make several of the features work. Embedding models are small compared to LLMs, but, depending on their capabilities, could be the memory driver.
> At first I thought it was the usual planned obsolescence. Then I realized it may be a true technical limitation.
It's shortsightedness on Apple's part, being both the first to the game, and the last to the game. Go back any number of years in the past decade. For at ;east 10 years Apple was telling everyone how great their local models are, and how amazing desktop-class their "neural" chips are. And yet here we are with them somehow incapable of running anything except on a few recent top-tier models.
The killer app would be a locally run Siri that learns about you and your preferences.
Llms don’t learn after they’re trained. At the most it could maintain a profile or some such. Which may be the equivalent of what you mean, but not as provocative.
> coming this fall
I believe we also heard that a couple years ago.
Went out of their way to show actual usage of these features on actual devices in actual people’s hands
Heh, I noticed the same thing, after the DaringFireball callout last year about the normal product demo progression. It looks "real" this time, but the question is how far along we are: will the journalists have a chance to play with it at the event?
It’s in the developer beta, people are posting videos of it, e.g.: https://x.com/iupdate/status/2064078761856037112?s=46
I noticed that, but I also noticed that it is ridiculously slow too which prompts the person to keep talking while the response was being generated.
After getting it in my hands, it's the same. At least 4 times slower for similar basic Siri responses. My guess is they are doing less local and more server-side generation to start as the on-device models might not be good enough yet.
In a pre-recorded video after who knows how many takes.
15 years ago they had the balls to run Siri live on stage: https://youtu.be/6rL9EL2LlrA?is=5yMQxs0C2VAC5Lwz
You could see jitter in the reflection on the MacBook as the guy typed, so that all looked great.
The responses came in very fast though, so I’m sceptical that the latency is representative (or that they didn’t cherry pick results, but they looked LLM generated). We shall see though.
Even for live shows they'd cherry pick specific scenarios that were known to work, and those would sometimes fail. IN a heavily produced pre-arranged pre-determined marketing video? It was as polished and made as smooth as possible.
I think that’s mostly okay, I just worry about the times. It was like 2-3 seconds to search and pull context then generate the answer.
I’m writing AI apps these days, and even pulling Gemini 3.5 flash on Google Cloud takes longer to get a multi-step response.
Obviously the video is not representative, and there are fast models on fast hardware. But if this takes 2 minutes it’s not very compelling to users.
Here's what I want: natural language interaction to achieve complex workflows in iPhone. Example: find the cheapest way to go from A to B and book it using the Deutsche Bahn Train app.
I still don't think Siri can do that ::angry::
The most notable thing here is that they finally have the primitives to make Siri actually useful across apps. I can't even use Siri to close Google Maps in my car right now.
for olders iphone models - i am on 15 - are they allowing any sort of extensions? that would allow some apps to make use of of to build things like better speech recognition etc..
i was waiting for siri update and bummed it is not supported on 15.
when i use my native language (i mostly do it when in carplay) to search songs etc.. it gets it wrong a lot of times.
+ a more integrated into things like imessages, whatsapp etc..
Why would they wait so long to launch this? With every passing day countless people further ingrain ChatGPT / Claude as their default chatbot...
It looks useful. But the demonstration this year is way below my expectations. They barely show any live demo, even few screenshots.
I wonder if Apple actually posttrained or at least finetuned this model or if it's just standard gemma. I feel it'd be bad practice if they didn't at least have some training atop it for apple's tools. Also you don't really hear much about apple's in-housed private compute servers anymore, did they get outmoded? I only hear about them using nvidia now.
Will I be convinced to change my iphone 6s? #suspense
Wait, I got an iPhone 15 Pro Max because supposedly it was compatible with Apple Intelligence. But now it’s not supporting this?
To be fair, the iPhone 15 line came out before Apple Intelligence was even announced.
It's the iPhone 16 line that feels a bit shitty not getting the latest and greatest since it was advertised as "built for Apple Intelligence"
Apparently the 17 Pro is the only currently released iPhone that will get the best local model. Which I suppose makes sense considering it has 12GB of RAM compared to the 16 Pro's 8GB.
Apple splits processing between an on-device and cloud-hosted model. As time goes on, devices will be more capable of doing more processing locally, and it would be expected that the cloud-hosted model gets more sophisticated.
Your 15 Pro Max supports Apple Intelligence. Newer phones can answer more questions without going to cloud infrastructure.
15 Pro Max is supported…
It should support everything except the new voice model.
Amazing how this time Apple found the `sweet spot` to release Siri AI when the letter combination A and I has fed up literally everyone.
Watching the keynote at our office on a big screen and everyone collectively sighing when they announced the name felt indicative haha.
I think it just feels uncreative? Siri as a brand has some value, but if you want it to feel like a watershed moment where old Siri is "behind us" finally, just give it a new name.
I don't think Apple has the option to rebrand Siri at this stage, assuming people actually call Siri by name. However, turning AI into 'Apple Intelligence' doesn't feel creative either.
> assuming people actually call Siri by name
This doesn't follow for me. They can trivially allow it to still respond to the old wakeword. They should absolutely change the name in the event they can finally make it useful, because "Siri" is (in my mind and many others') a synonym for "hapless idiot." "Thanks, Siri" has been uttered hundreds of time in my house and my car, and 100% of the time it's sarcastic.
Same thing in my head. I only see this going one way, which is tons of people hear that Siri “got better” after this update.
Many of those people will speak a language that’s not English, or live in the EU or China where it’ll still be “Siri”, not “Siri AI”.
“Do you have the new Siri?”
“Yeah I updated… but she still seems so dumb”
“Oh yeah… well that’s Siri for you I guess”
Horrifying for marketing folk, I would presume. You’re just setting people up to confirm that Siri is always useless and improvements are invisible.
we were doing the same thing and giggling a bit that it's basically "AI AI" now. realistically a lot of people thought of Siri as AI already.
Or just keep calling it Siri, and announce "hey look, Siri does some cool new things."
AI is a technology, not a product. Consumers don't care about technologies, they care about what the product does versus what they currently have.
I think Jobs was an asshole, but one good thing I can say about him is that he understood the difference between technology and products. Imagine if they had called it the "iPod HDD."
The ergonomics of “the new X” sort of fall apart when you’re releasing it in stages. (Not in EU/only in English) It spawns a lot of conversations like “do you have the new Siri? Uh… I think so? It’s still crap though.” You cart around this bad brand image because you pitch this big watershed moment and 2/3rds of people are still using the “wrong” Siri.
Siri and Voice Control were both usable during the same time and it feels like it could work here too.
Totally agree that AI is just an implementation detail though. IMO that new product name should NOT have “AI” in it at all.
It's going to be weird though when my phone has Siri AI and my Homepod has Siri... "please ask on your iPhone" edition. I also don't quite get the distinction of Siri as an app versus the Siri I yell at to make my TV do something.
Amazing how someone again finds a meaningless thing Apple does better than the rest then blows it out of proportions. Makes you wonder if they are on Apple's PR team.
So you noticed the five ecosystems being shown working together, not perfectly but better than everyone else.
It looks bad from every perspective. I've never seen two apple's in one URL for product category before. apple.com/apple-intelligence
To prove my point, I opened a random date on the Apple website matching today's date to compare. 16 years ago, June 8 (1) Apple released the iPhone 4. There's still no room for jokes about that release, and from this perspective, calling their AI 'Apple Intelligence' feels weak compared to what they used to deliver.
I agree that some years ago Apple was the strongest in marketing, their brand team had been setting the bar for tech, but I simply can't say that anymore.
1. https://web.archive.org/web/20100608073904/http://www.apple....
Apple notably threw shade at the existing AI implementations, with an emphasis on making Siri AI more human-focused.
The stock price definitely didn't like it though.
The stock market is notorious for dropping on almost any Apple conference or announcement.
I mean, seriously, AI = Apple Intelligence?
It's not even funny, it's not smart. It's like if they released MS Siri and said it's Mac System Siri.
> I mean, seriously, AI = Apple Intelligence?
For pedantry's sake, they were saying "AI = Apple Intelligence" last year as well, so it's not like they just pulled it out of their butts now that popular opinion has turned against AI.
Exactly, I thought one year was enough to prove that everyone reads those letters differently nowadays.
I feel like the hate came more from the "Available today for Developers and later in Beta" than anything
I’m honestly surprised Apple didn’t retire the Siri brand.
At this point, “Siri” has a pretty strong cultural association with being underwhelming or unhelpful. Even if the new version is dramatically better, convincing people to give Siri another shot may be harder than launching the same technology under a new name.
Feels like a missed opportunity to reset expectations.
Yeah missed opportunity. They could have even had a fake funeral for Siri like Jobs did for OS 9, or a "retirement party" or something. Leave the Siri brand behind and launch this as something brand new.
The automatically update of my compromised passwords on websites is very impressive, and I wonder how it's achieved.
I don't care much about Siri, and not a lot about Apple (other than as an investment), but Apple is generally really good about putting out polished tech, and so I'm curious if Siri AI will be up to their usual standards, because if so, it represents a significant usage of AI that has solved hallucination issues.
But that's a big If!
With the amount of people that have iphones in this world, I am more than sure that somebody will be able to make it do something wrong.
Apple knows this which is why it is taking years to test and iron out the kinks. But somebody somewhere will make it hack a social media account, or give over somebody elses credentials, or generate illegal child images etc.
The big if is does it work on device without phoning home? For example, Google or Meta as a Ad/data collection company wants to phone home for everything because they want to collect data on you, what will this Apple solution do?
There's extensive writing about Apple's Private Cloud Compute architecture.
It's beyond my expertise, but it's publicly available if you're curious.
> I'm curious if Siri AI will be up to their usual standards
You clearly never used Siri before
Ready later this year... What I want from the phone are low-level APIs exposed to an agent I can talk/type to without lag. I wonder how long it will take or if a new AI phone will come before these features. I would love Siri working because Apple's design and hardware are good.
I sincerely hope this can all be disabled.
'later this year' lol uh ya, ok apple (only had 3-5 already to make it even a 1-10 usable)
I wonder if they got around to the Finder date-sorting bug that's been around for about 10 years.
What’s the bug
Associated post: https://www.apple.com/newsroom/2026/06/apple-introduces-siri...
The demo video there is so underwhelming. They show very basic stuff, which I assumed Siri was already capable of doing... not sure what the big improvement is
Based on my experience, I assume Siri is not capable of anything more than setting timers or referring me to my phone to see the search results.
And that’s pretty much what they show in the demo, plus asking it for directions on Apple Maps (which it can also do already), and searching for pictures on the Photos app (which I just tried and it can’t do - so it looks like that’s the main feature)
If it's going to be anything like Gemini on Google Pixel, it'll be great at everything except for trivial tasks like setting timers :).
The one thing I've been trying to figure out / hoping will get a fix is that in the apple intelligence settings panel there is an 'extension' that allows it to use chatgpt. I would like to be able to have an extension for local models and/or custom apis.
We heard a lot of things, but unfortunately, I'm pretty sure it won't work as expected.
The screenshot about pho is funny to me. Bean sprouts are not a good source of fiber. Noodles are not especially healthy. The broth base is not fish sauce, nor is fish sauce where broth gets most of its sodium. Slop city!
Yeah, "Fiber: good source" when 100g of raw bean sprouts gives you 1.8g fiber (less than a 2" kiwi), and pho comes with much less than 100g of sprouts.
Pho is a pretty bad source of fiber.
It sucks that we're skipping over such good tools like cronometer.com to figure out what we're actually eating and going straight to hallucination, adding more confusion to nutrition.
Seriously - that was bizarre. Why would they not choose a healthier meal or something more unknown?
Apple's execution on AI is the worst of anything I can think of they have worked on in the past 20 years. It's embarrassing they announced this with a vague "coming this fall" when they basically have completely lost credibility in their ability to ship AI features considering it was initially announced YEARS ago.
I think a lot of it is the old "perfect is the enemy of good" with Apple trying multiple times now to announce this big basket of all these AI features supposedly coming all at once instead of just regularly shipping new useful AI integrations every month. There was so much easy useful shit that was immediately apparent as soon at OpenAI dropped that first big voice mode years ago coupled with basic app integrations. Particularly in the context of the AI labs that are operating in that lane almost too much where it seems a new model or mode comes out every two weeks.
> Create useful shortcuts with ease, update passwords with just a tap, and find relevant emails faster than ever
Can't wait for unexpected password updates and naughty mails accidentally sent to my boss...
Please don’t suck.
Right? I've been waiting a lot of years for an upgrade to my voice-activated timer setter/music player launcher.
Can't even get that right sometimes. A few weeks ago I somehow accidentally activated Siri and it decided that what I wanted was for it to play some kind of terrifying industrial electronic noise music that scared my kid.
The only thing I want to know about this new Siri is how to turn it entirely off.
From an sysadmin perspective, assuming Apple provides full MDM toggles you can typically toggle these off one way or another. (.plist / preference keys, etc).
With the most recent 'Apple Intelligence' function, it took a while for Apple to grant the ability to disable/enable each feature, then a bit of time for the respective MDM Software developers (Jamf, etc) to provide toggles.
Sounds good, where’s the “just kill this” toggle in my iPhone settings? I ain’t gonna bother with mdm etc etc I just need the damn thing to work and do what I need.
You could just not use it.
Oh I’m definitely not going to use it. But apple tends to try to put these things in your face and get you to use them (seen all those purple buttons in Pages lately?). A global “just don’t show me or try to get me to use this crap at all” switch would be very welcome. Bonus if it also frees up however many GB of space this garbage will take whether I use it or not.
You don’t know what’s in your face until you use Copilot.
No worries I would never touch anything Microsoft with a 10-foot pole.
From your comments it sounds like Apple are just peachy as well.
I’m mad at Apple for dropping the ball with the sloppy mess that is iOS 26. I trust they’ll see the light and fix it. Miles of difference with Microsoft - I’d sooner renounce computers entirely and quit my job than be in a position where I have to use Microsoft stuff regularly.
I agree, I am kinda disappointed by how much AI there is in this update, I was hoping for something different and exciting.
I think it's important to note what they didn't talk about. They briefly scrolled a long list of performance improvements and mentioned a few. Personally, I am very hopeful that the fact that they only talked about these high level AI features means the other engineers got to spend time focusing on performance and quality.
Fixing the shit show that iOS 26 is and making it so my shiny new iPhone 17 doesn’t visibly struggle rendering simple ui elements should be considered table stakes for iOS 27, I wouldn’t expect them to mention that too much as it’s like the bare minimum we expect for the update.
If that Siri orb fails to respond after this release, I'm done with Apple.
the same :'( what i really expected from Siri,
- Siri please suggest an organization for this folder - Siri over my last work in this app can summarize what i am struggle ?
Pro active:
- Hey, the last hour you exchange 30 mails from the same subject, i look with your team ai and all have same struggle, based your key points in communication is X,Y,Z, this an mail for final align
- Your and your partner don't have quality time in last day, i see has the seat available in your favorites restaurant for next hour do you want made an appointment ?
There's only one thing that matters, that Siri would become an independent app that can see through its calling histories, which should have been done since the first day it launched.
Here's hoping they've finally fixed iOS's terrible dictation.
They actually released a pretty good dictation model in the OS 26 updates. Someone made a cli utility to play with it:
https://github.com/finnvoor/yap
I tried it and was pretty impressed. That said I haven't heard anything yet about them switching to this for the text input voice dictation in iOS but it would be really nice.
They better make it really good. Most kids get their facts from it. One has to argue with their kids when it gives wrong information.
87% US teenagers own an iPhone. ~35% teens own an Apple Watch.
> Private Cloud Compute
> Your data is never stored
> Used only for your requests
> Verifiable privacy promise
Apple is cooking. Although at that point might as well bring the cloud features to more devices. Yeah it costs more but also locks users in harder.
Ant they still failed to get past EU privacy regulations at launch.
It’s weird it says I can ask Siri about a document in front of me but can I ask it about a webpage I’m currently reading?
(It’s been driving me crazy there’s no “AI this” button to discuss whatever is on my screen.)
Edge has had this for a long time. I can highlight a string and right-click, 'Send to Copilot' and click "explain" and it'll prompt it to 'explain this passage, particularly in the context of the current page.'
Note: I have MS 365 personal or whatever it's called this week so I'm not sure how Copilot acts for a completely free user.
I'm guessing they'll integrate with the double-tap-the-bottom-of-the-screen feature that pulls up siri in front of a screenshot. Currently it doesn't seem to hook into "visual intelligence", and needs to call out to ChatGPT to do anything with the screen contents.
Tangential note on
> double-tap-the-bottom-of-the-screen feature that pulls up siri
It’s disabled if not using Apple Intelligence, and can’t tap screen while talking to Siri (it dismisses instead).
Now they’re gating features to the M3 I’m not convinced wouldn’t work on expensive Apple Silicon predecessors… am more convinced the double tap disable is intentional.
Today's presentation was about the things you can do in the next release. And the asnwer to your question is "yes", in Golden Gate.
That would be a finder feature vs a safari feature? They talked about safari’s capabilities in a separate segment.
They are very far behind, this doesn't feel like catch up.
Hopefully it’s an app that can be uninstalled on macOS, where I use almost none of Apple’s own apps, yet are prevented from removing them (unlike on iOS).
How come Apple has become so slow in shopping compared to equivalent companies like Google and Microsoft
Apple Shortcuts have felt like a blatantly obvious AI play to me for a while now.
The interface for creating them manually has been so bad for so long, it feels clear to me that LLM-driven shortcut orchestration was always the endgame. Apple built up their ecosystem of composable "tools", and then trained an LLM on how to call them.
The result, IMO, is the first OpenClaw/Hermes competitor that's feasible for use by the general public.
Everyone with a paid Claude or ChatGPT that they're struggling to use to the fullest is going to have very little reason not to swap over to an upgraded iCloud+ plan (if they don't already have one). I suspect we're going to see mass cancellation of $20/mo plans very soon.
OpenAI's timing for removing their temporary increased usage limits is looking pretty unfortunate...
Why do I need shortcuts though, I want that to be transparent.
Good point, that's probably gonna be the hardest sell.
I have shortcuts set up to count the hours I log in my work Google calendar and copy them to my clipboard to help me prepare invoices.
So while I've already been sold on what Shortcuts can do, getting the general public to see the possibilities is probably gonna be a challenge.
I think Shortcuts has a few massive flaws that would give me pause enshrining it as middleware for an important thing like a "mainstream OpenClaw".
1 is performance. It's slow. You can run one within the app and literally watch execution flow from one block to the next. Absurd, for the CPU power at hand.
2 is reliance on developers to deliberately implement hooks and "intents" when the developers of at least half of apps including most "big company" apps do not care to bother, often because 95% of their app's surface is actually cross-platform stuff.
Example: There are no shortcut actions for Google Calendar, and Gmail only has one real one which is a generic send email. No "search email" etc.
I'd rather see Apple lean into "computer use" to allow it to use any app that displays things on the screen, but IDK how you make that safe.
Hm, second try? And Siri AI again without dates. First time it was also “later” but was postponed for how many years?
No mention of making my homepod mini “Truly helpful”, so I’ll stick to just sticking timers, I guess.
__Truly helpful__ it's funny to think about reasons people put these words there
The ability to tap into on-device foundation models within apps is pretty baller. The rest is banal
Does this not work with HomePod? How is an AI voice assistant not compatible with the smart speaker?
That old hardware can't handle it. HomePod, the Apple TV, and the watch will all get updates in the next few months to streamline this.
A multi trillion dollar company not able to create open apis for competition … definitely the EUs fault
Siri AI is a mouthful… they should have kept it simple and stuck with Siri
How is this different from the chatgpt apple intelligence thing from last year?
This time they really promise it'll ship
But fundamentally, the real difference is they have now bought and white-labeled Gemini to replace all the stuff they failed to make 2 years ago.
We need Siri API. Features like this are nonsensical in an era where I can wish up any feature or workflow for an app.
People outside HN will begin to expect they can do anything with a computer in the same way they expect to be able to say anything in a support chat. Using pre-LLM automated chat feels like a joke. You enter the chat expecting to having a conversation and instead you get a GUI phone tree.
This is exactly how it feels to use any of the AI tous from Big Tech and others.
We have entered the era of deeply personal computing. There are so many incredibly personal features that no mega corp could ever be expected to build. Now that lay people can build things, let them!
There's a podcast that I listen to which is translated to a bunch of different languages. It's exclusively on Spotify with no RSS feed. I have a cron that checks daily passes it to an LLM and notifies me as necessary. I did not code a line. I only set up an OAuth endpoint.
Enabling your customers to do things like this will make them incredibly sticky too! So please for the love of God, let me glue Siri into whatever I want.
Its wild to me that people use those apple emoji people. It looks so bad.
That’s your clue that it’s AI, or do you want it so realistic that you don’t know the difference?
Sorry, i've not had an apply device in ages. Is the AI using the emoji icon or are people using it? I assumed people were using it.
This will come very soon and be the default, so it doesnt really matter what people want in this area.
Does the UK get this new AI and not the EU?
If this is good, I might finally ditch my 12 mini.
Great, as long as I can switch it off and use my phone as I always have, I'm happy for them.
I can't wait to take a photo of a cricket ball and ask it what it is, ffs.
These people need to get out, touch grass, watch trees swaying in the breeze, and put their phones down before they lose toonmany neurons.
[dead]
How many companies did they just Sherlock?
They are a few years late to the party
pardon me but I'm not getting it, what's new or exciting about it (supposedly)?
Honestly I don’t have much faith in Apple intelligence when it can’t even search my settings.
My number one question is can I turn this shit off? I don’t want AI infecting and having access to my whole fucking life. They don’t say anything on this marketing page.
Week if its like the current apple intelligence feature,I don't care if we don't get in the eu.
More or less stuck AI in all the obvious spots, which will probably be fine I guess. Not super exciting!
Can Apple now focus on rolling back Liquid Glass “upgrade”?
If you watched the keynote they added transparency controls, adjusted the refraction, removed the different window corner radii into one, and cleaned up sidebars and toolbars. They also updated icons to have more depth to them so you get a harder border on key elements.
I still look at older MacOS screenshots and think a lot of it looks better, but directionally they are improving Liquid Glass.
> We’re deeply disappointed that our EU users won’t have Siri AI on iPhone or iPad when we share our new software releases later this year
I really enjoy this lag. Apple with the whole DMA made iPhone completely dull to my eyes. Previously? Updated yearly. Now? 3+ years without replacement and probably will stick to it for next 2-3 years.
Sure maybe in US Apple is fun. But in EU it's.. boring (and not like a Golang boring, just boring)
It's bananas to me that people want to put LLMs in critical software paths when the 2nd most valuable company on Earth can't do more than the most basic shit with them. Like, look at this list of actions:
- Take a selfie
- Create a reminder
- Call Vicki
- Rotate photo left
- Create a new event (do you create an old event?)
- Send an email
- Resume my podcast
- Create a note
- Add photo to album
...can't iOS do literally every one of those things already? What the fuck is happening?
This is disappointing. I had hoped when Apple revisited AI that they would lean into agents more and give us some sort of agent interface between the phone and a model running locally on your Mac at home. More niche for sure, but much more powerful. Instead we're getting more generic AI tie-ins to apps and "suggestions".
Seriously - this would be a major cool thing to actually use AI for. Run some local AI basic stuff on phone and have it securely connect to your own home device for more advanced tasks or control your home agents. Could even reduce their own need to host cloud AI compute
For real this time...
missed opportunity to call it "VibeSiri"
sloppy mc-slop. dogshit nobody needs. AI is innovating horizontally, like web frameworks.
Hey look! Here's something new that we could already do but now it costs more and takes more engineering and.. AI
[flagged]
[dead]
[flagged]
[dead]
[dead]
[dead]
[flagged]
Interesting. I'm quite bearish on AI in general, but this will get me to upgrade my iPhone 14 Pro to whatever they release in September.
Why?
For me, it feels like a prompt injection based security nightmare.
Literally every file on my mac and every site I browse is potential malware.
Edit to add: every email and text message as well.
[flagged]
Apple's "New Coke" moment?
all the limitations of on-device with none of the benefits, it seems? they gotta get SOMETHING out there and soon but idk, i would probably feel safer running a chinese model through a 3P iOS app shell vs. trusting Geminiri to not snitch if i cared about the sanctity of my personal information.
> all the limitations of on-device with none of the benefits, it seems?
What do you mean?
i think i just mean if they are using google models, then i believe every query is going to google no matter what apple claims about "keeping it local." whether google does anything with it, separate question i guess, but i imagine it will ALSO be slow to simulate the protection apple is selling. and sure, it's a catty comment, i'll own that. but that is my read on the announcements and demonstrations.
I go on long walks and talk to ChatGPT in depth in its conversation mode about programming and computing in depth.
That’s what I expected from Siri but you can get in from ChatGPT .
Newest phones get latest models
Genius way to sell more phones
Really they are just selling on device Ai
I like the idea for normal people. Day to day usage who ignore hallucinations is a big market.
> Siri AI coming in English later this year.
Strange way to phrase it, but okay.
> Siri AI will be available In beta later this year and requires an Apple Intelligence–enabled device set to a supported language. Available in English to start. Siri AI will not initially be available in the EU on iOS and iPadOS.
Ah okay, not EU enabled. The only reason for this, in my tinfoil hat, must be for data farming.
I was at Best Buy in the Apple section, and I asked its AI "What is the best value in four year old MacBook pros." It pointed me directly at an over-under washer dryer for $3400. Quite obviously, it was trained, like a drooling puppy by Madison Ave.
Wait... don't tell me... there is an App for that.