Siri AI
660
FR version is available. Content is displayed in original English for accuracy.
Advertisement
Advertisement
⚡ Community Insights
Discussion Sentiment
66% Positive
Analyzed from 18712 words in the discussion.
Trending Topics
#apple#siri#don#more#app#iphone#phone#years#data#things
Discussion (684 Comments)Read Original on HackerNews
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".
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.
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.
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.
They don't even seem to get the basics right, why would I want another layer on top?
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.
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.
Because the problem now took a whole afternoon to solved and sapped your creative energy instead.
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.
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.
"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 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 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.
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)
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.
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.
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.
https://www.normaltech.ai/p/new-paper-towards-a-science-of-a...
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.
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.
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).
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.
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"
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?
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....
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.
Have a conversation with the average Ai power user (outside of tech / coding) and this is the level the conversation will be on.
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.
It's just not compelling to say that an AI can do an easy task quickly. This is still worth zero dollars to me.
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"?
It sucks. This is why none of my local supermarkets have real checkouts anymore.
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.
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.
...
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.
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.
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)
https://livebench.ai/#/?highunseenbias=true
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.
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.
"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.
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.
Orders of magnitude less than the literal trillions that others have?
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.
ref: https://www.evehome.com/en/eve-blinds-collection
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.
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.
What do you mean exactly? Audio conversation only? If so I don't see it very practical for most of the things
I think that we are now in the realm of diminishing returns regarding these chat assistants.
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.
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).
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.
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.”
https://clarksonlawfirm.com/lp/apple-intelligence-false-adve...
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
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.
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.
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.
Facebook got roasted for this, but now the EU wants the same open data policy from every big tech company.
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.
>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...
Apple’s performative DMA outrage is getting more pathetic by the iOS version.
Apple cares greatly about their brand yet this has hurt their brand like nothing else in the past decade
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
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.
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.
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
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?
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...
Why? What strengths and structural advantages do you think thy have?
What black swan situation could arise that Apple cannot counter?
If there’s truly an existential threat to its device business, Copy Well
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?
So while they could win, it’s pretty hard to get hyped about it before we see real-world tests.
Screenshotted in case they change it https://imgur.com/a/n1I3z8g
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.
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.
AppleScript even had "dictionaries" declaring their commands and everything, would have been perfect way teach LLMs how to automate applications.
"Try describing something different for the shortcut."
I guess I shouldn't be surprised that it still doesn't work.
Seems like the logical next step
Siri seems to rarely get better and sometimes actually get worse.
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…
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.
Can you give a couple of examples?
This has been a problem on iOS since the dawn of time and has nothing to do with AI
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.
>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.
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.
git + somesite.com.gpg
https://github.com/FiloSottile/passage (or: forked using AGE instead of GPG)
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.
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.
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.
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.
I don't really believe in Apple being that quality team.
Why?
They have no expertise in this area and their software quality as never been worse.
AI could potentially help solve those unpopular site/app/whatever edgecase.
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.
E.g. "the user asks if their Bitcoin private key is unique, let's make a web search".
Apple cannot compete in AI and has to adopt Gemini
Google is a really amazing 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.
> Aga: have you heard of Calanthea? It’s a plant.
Really groundbreaking use of AI!
wow! we are in the future!
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.
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.
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
If OpenAI makes their own AI-phone, do they have to let Anthropic and Deepseek run their models on it too?
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.
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.
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
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.
I'm sure you guys have tons of others to list.
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.
I'm almost sure that sometimes searching the same thing will give you the result and sometimes it won't.
I'm sure they customized some of it, but this looks basically like Gemini integrated with iCloud instead of Google Workspace.
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.
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.
Literally every file on my mac and every site I browse is potential malware.
Edit to add: every email and text message as well.
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.
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).
Apple’s entire software stack has a branding problem.
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.
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.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.
It just reads like arrogant foreigners throwing a tantrum over our 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.
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.
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.
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?
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.
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.
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.
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.
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.
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.
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.
Because we applied the learnings from computers and phones have ended up way more personal?
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.
2 is permission to act, but you might want to deny access to some parts such as sending email and scope to summarization
What is the purpose of that?
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..
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.
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.
> 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 :(
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).
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.
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.
I believe we also heard that a couple years ago.
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.
15 years ago they had the balls to run Siri live on stage: https://youtu.be/6rL9EL2LlrA?is=5yMQxs0C2VAC5Lwz
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.
Can't wait for unexpected password updates and naughty mails accidentally sent to my boss...
- 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 ?
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.
Your 15 Pro Max supports Apple Intelligence. Newer phones can answer more questions without going to cloud infrastructure.
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.
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.
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."
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.
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....
The stock price definitely didn't like it though.
It's not even funny, it's not smart. It's like if they released MS Siri and said it's Mac System Siri.
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.
But that's a big If!
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.
It's beyond my expertise, but it's publicly available if you're curious.
You clearly never used Siri before
87% US teenagers own an iPhone. ~35% teens own an Apple Watch.
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.
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.
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.
> 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.
(It’s been driving me crazy there’s no “AI this” button to discuss whatever is on my screen.)
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.
> 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.
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.
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.
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 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.
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.
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.
- 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?
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.
I still look at older MacOS screenshots and think a lot of it looks better, but directionally they are improving Liquid Glass.
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)
I still don't think Siri can do that ::angry::
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.
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.
That’s what I expected from Siri but you can get in from ChatGPT .
Wait... don't tell me... there is an App for that.
What do you mean?
Genius way to sell more phones
Really they are just selling on device Ai
> 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.