When my phone is sitting on my countertop and I say "hey siri, add eggs to my shopping list" I expect it to add eggs to my shopping list, as it has done so for years, instead of saying "you'll need to unlock your iPhone first."
When I'm driving my car merging onto the highway and I see furniture on the ramp, and I say "hey siri, call the local non emergency police department number" so that I can report it, I expect it to say something like "I have the [city name] non-emergency phone number, want me to call now?" Instead of it saying "calling the New Jersey police department" which is halfway across the continent from me, and then giving me no way to hang up the phone hands free forcing me to grab my phone from the console and quickly try to hang it up while I'm driving at highway speeds.
- When I'm driving and I say "Hey Siri, navigate to Walmart" I expect it to begin navigating and not show me a button which I must press to begin navigating.
- When an alarm is going off on my phone and I say "Hey Siri, turn off the alarm," my laptop or HomePod should not answer to say there is no alarm.
- When I say "Hey Siri, add eggs to my grocery list," my HomePod should not think for 20 seconds and then tell me to check my connection, despite both it and my phone being on very strong wifi.
- When I say "Hey Siri, add eggs to my shopping list," it should add them to my "Groceries" list and not tell me there is no list named "shopping."
Yeah, the “new” Siri described in this article is not the Siri I’ve been interacting with. Basic speech to text is still flawed; I have to enunciate slowly and clearly with no background noise or it will make absurd mistakes that would be obvious if the context of the prompt and the phone were considered. Adding a reminder to a list is now slower and more error prone than with the old Siri, like you said. I have to sit and delicately hold the phone or watch to avoid disturbing it while it thinks, lest it decide that I’ve moved on and give up on its task. I tried to push it a little by asking for the title of the last podcast episode I listened to; it started talking about a podcast I’d never heard of nor even pulled up on the phone before garbling the second half of its sentence and then going silent.
I suspect it's regressed because they actually don't have good e2e evals comparing the two experiences to test the regression and UX. Like the simplest evals you want is just a shit ton of STTs across human speech accounting for dialects, accents etc and then figuring out how the distribution of errors has moved or increased on the axis you care about. That they likely have. But then you have to integrate that into the product to capture the entire funnel of errors, which I suspect is the part they don't have. The stacks are not "vertically integrated" to test quickly.
Dictation being so terrible is truly wild to me and I don't understand it. On my M5 MBA literally any of the vibe coded local dictation apps that come up (and there are like, hundreds of them!) are better than the built in macOS dictation. They are much faster and the AI text cleanup is so much better than whatever the Apple foundation models are doing.
Another example from yesterday: I’m using CarPlay. A text message notification pops up. I ask Siri to read the message. Siri says there are no recent messages. More notifications pop up. Siri still has no idea.
> Instead of it saying "calling the New Jersey police department" which is halfway across the continent
Yeah, that's Apple's generally broken maps search. It regularly and consistently just brings up the wildest things from the farthest reaches of the Universe (well, usually restricted to Planet Earth, barely), when I am clearly looking for something local.
> Even today, Federighi carries a reputation for having been an AI skeptic, concerned about the technology’s privacy, security, societal and employment implications.
I’m kinda in agreement with Federighi on these concerns. Though he hasn’t delivered on software quality (from my perspective), I believe it’s good to have someone who keeps these concerns in mind (not necessarily SVP of software).
I tried Siri AI on iPhone recently thinking that it’d be great. I found that Siri AI can’t even navigate to a given URL to get look up information (the response was, “I can’t access specific web pages”). At this point, it seems like Siri AI is probably good for accessing/working with some data on the iDevice and/or doing research on the web.
Siri AI to me when I asked what it is even good for:
“I can help with a wide variety of tasks on your device.
I can help you draft emails and messages, set reminders, create calendar events, and find information from the web. I can also help you organize text, perform calculations, and manage your smart home accessories.”
The software chief who pushed out the previous AI chief (who wasn’t enthused by LLMs) must move out for Apple’s AI…and even its own consumer release software…to improve.
What I don't understand is how I can, allegedly (I've never tried), ask siri to dig through my photos and iMessages to look for stuff, but when I ask it to summarize the website on the screen it says it can only see literally what's on the screen and can't scroll down or grab the entire article that's already been loaded.
If Apple forces developers to support Shortcuts, ensures true Mac-to-iPhone parity, and lets users generate scripts using natural language AI, it’s game over. Nailing this native integration would single-handedly kill the market for almost every derivative AI app out there.
Interesting notes about Craig's management style in that article. I have worked with a few C-Level folks with similar traits (e.g. "smartest in the room", etc.) and never seen great outcomes. Have been hoping for years that MacOS would get back to being a delightful upgrade -- faster, better, etc. For me, that has been lost during the Craig decade or so. Note: This is just an opinion, and not well informed apart from seeing a guy like guys I worked with. I.e. not Craig's fault, and it might have been worse if not for him.
I have an iPhone 16 which supposedly was made for Apple Intelligence. I enabled Siri intelligence beta and all requests slowed to a halt, even calling someone. I’m guessing that everything was being processed in the cloud as opposed to my device. It couldn’t even detect that I had a timer running that it started.
So I disabled it and rely on normal Siri again because somehow it’s better.
It's a crowded field and Apple has almost no expertise in AI, but hopefully they can find some angle that makes use of their massive locked-in customer base and huge piles of cash.
I'm not sure about "no" expertise in AI. A lot of their AI work has been in efficiency and getting models running on the fanless rectangles everyone has in their pockets. Same for Google. They just don't seem to be interested in developing big, expensive frontier models. They've leaked CLAUDE.md in software releases a lot, so for internal development it seems they're fine paying Claude API prices while Anthropic burns piles of cash on training and advancement.
What Apple needs to accomplish with AI, given their lack of a home-grown frontier model, hasn’t been done, at scale and with refinement, by any company anywhere, except maybe Perplexity. They’re struggling to refine a unified consumer AI platform that is built as an abstraction layer.
I'm now using Grokbot and a custom written equivalent regularly in my day to day, in fact I'm so enamoured with this way of working that I'm actively building it into most of my new projects and tools... meanwhile I can't get Siri to respond to me reliably any more. This whole space is moving so fast a yearly update cycle isn't going to be enough.
That said, I could see there being a free Siri, and a paid for version that is far more powerful. Add it into Apple One? It's also worth noting here that I'm over caring whether this runs on device, I care more than it works.
Rather than making posts/threats about how they're going to ramp up their permissions systems even further... instead they should be scrambling to turn a conversation with your phone a first class interface, otherwise people will discover tools that live outside the Apple ecosystem and then people will gradually move away from it (I can't see them actually opening it up unfortunately). This will quickly become an existential threat to them, because as great as the iPhone is.. if it doesn't fully work with a user's AI agent(s) it will be a problem for Apple.
As much as a legend this guy is, it's the call of the CEO to "retire" him. He would have pursued a more ambitious ai strategy if he really wanted to, so just urging him to do that again won't cut it.
They're in an interesting spot with AI, any rival has to run behind Siri to do the same things but they're not guaranteed a 30% piece of their revenue anymore, they're leaning on Google for datacenters so there's a real cost for usage but they're not yet worth paying to use.
> better integrated with its technology than the competition
And even this is looking more like short-term relief than a lasting advantage, and in a lot of countries they've already exhausted that option.
I would like at least the basic things to work. Like spelling/auto-correct on my iPhone. You'd think after 10+ years of using the same apple account across many devices, it could learn to spell the word I've used 100 times already? It looks worse than the simplest markov chain. Not to mention finding files in Finder - impossible. I tend to instead use find/grep from the terminal, and worst case, when I can't remember exactly the name (but have some rough ideas of what it may be called), I get Codex to find it for me.
It is quite difficult for me to think of a mass-consumer AI business here. You can add a hundred little features scattered throughout iOS that give real improvements, and maybe put that behind a subscription?
> Another concern is Federighi’s preference for planning over relatively short periods, typically 12 to 18 months, compared with the multiyear road maps favored by Apple’s hardware organization.
I feel like everything I read about Apple's AI strategy is full of these kind of "huh?" moments. How can someone who wants Apple to pursue an "Ambitious AI Strategy" simultaneously believe that the software org can produce meaningful multi-year roadmaps right now?
The future of software and technology was already looking pretty bleak pre-AI. Now the AI slopification of everything feels like the final death blow to me.
Really? AI is finding and fixing critical bugs that have existed for decades. Hobbyists are recreating and remixing closed source games that have been abandoned. It has never before easier for a total amateur or an experienced coder to create as capable tools as now.
It is very disappointing what Apple has done in the past two years with Siri and AI.
First they did an ad for Apple Intelligence and Siri. Then two years silence. Then they launched it, it’s worse than it was before, not available in the EU and they use models from other companies.
They're a hardware company that made trillions of dollars selling hardware. "Commoditize your Compliment" is one of the oldest business strategies out there.
A hardware company whose strongest gross margins come from services and has been trending in that direction for over a decade. Heck, if trends continue, their margins and profits might be predominantly services based in the not to distant future.
Siri needs to be useful, period.
When my phone is sitting on my countertop and I say "hey siri, add eggs to my shopping list" I expect it to add eggs to my shopping list, as it has done so for years, instead of saying "you'll need to unlock your iPhone first."
When I'm driving my car merging onto the highway and I see furniture on the ramp, and I say "hey siri, call the local non emergency police department number" so that I can report it, I expect it to say something like "I have the [city name] non-emergency phone number, want me to call now?" Instead of it saying "calling the New Jersey police department" which is halfway across the continent from me, and then giving me no way to hang up the phone hands free forcing me to grab my phone from the console and quickly try to hang it up while I'm driving at highway speeds.
- When I'm driving and I say "Hey Siri, navigate to Walmart" I expect it to begin navigating and not show me a button which I must press to begin navigating.
- When an alarm is going off on my phone and I say "Hey Siri, turn off the alarm," my laptop or HomePod should not answer to say there is no alarm.
- When I say "Hey Siri, add eggs to my grocery list," my HomePod should not think for 20 seconds and then tell me to check my connection, despite both it and my phone being on very strong wifi.
- When I say "Hey Siri, add eggs to my shopping list," it should add them to my "Groceries" list and not tell me there is no list named "shopping."
Yeah, the “new” Siri described in this article is not the Siri I’ve been interacting with. Basic speech to text is still flawed; I have to enunciate slowly and clearly with no background noise or it will make absurd mistakes that would be obvious if the context of the prompt and the phone were considered. Adding a reminder to a list is now slower and more error prone than with the old Siri, like you said. I have to sit and delicately hold the phone or watch to avoid disturbing it while it thinks, lest it decide that I’ve moved on and give up on its task. I tried to push it a little by asking for the title of the last podcast episode I listened to; it started talking about a podcast I’d never heard of nor even pulled up on the phone before garbling the second half of its sentence and then going silent.
I suspect it's regressed because they actually don't have good e2e evals comparing the two experiences to test the regression and UX. Like the simplest evals you want is just a shit ton of STTs across human speech accounting for dialects, accents etc and then figuring out how the distribution of errors has moved or increased on the axis you care about. That they likely have. But then you have to integrate that into the product to capture the entire funnel of errors, which I suspect is the part they don't have. The stacks are not "vertically integrated" to test quickly.
Dictation being so terrible is truly wild to me and I don't understand it. On my M5 MBA literally any of the vibe coded local dictation apps that come up (and there are like, hundreds of them!) are better than the built in macOS dictation. They are much faster and the AI text cleanup is so much better than whatever the Apple foundation models are doing.
Another example from yesterday: I’m using CarPlay. A text message notification pops up. I ask Siri to read the message. Siri says there are no recent messages. More notifications pop up. Siri still has no idea.
This used to work just fine.
> Instead of it saying "calling the New Jersey police department" which is halfway across the continent
Yeah, that's Apple's generally broken maps search. It regularly and consistently just brings up the wildest things from the farthest reaches of the Universe (well, usually restricted to Planet Earth, barely), when I am clearly looking for something local.
> Even today, Federighi carries a reputation for having been an AI skeptic, concerned about the technology’s privacy, security, societal and employment implications.
I’m kinda in agreement with Federighi on these concerns. Though he hasn’t delivered on software quality (from my perspective), I believe it’s good to have someone who keeps these concerns in mind (not necessarily SVP of software).
I tried Siri AI on iPhone recently thinking that it’d be great. I found that Siri AI can’t even navigate to a given URL to get look up information (the response was, “I can’t access specific web pages”). At this point, it seems like Siri AI is probably good for accessing/working with some data on the iDevice and/or doing research on the web.
Siri AI to me when I asked what it is even good for:
“I can help with a wide variety of tasks on your device. I can help you draft emails and messages, set reminders, create calendar events, and find information from the web. I can also help you organize text, perform calculations, and manage your smart home accessories.”
The software chief who pushed out the previous AI chief (who wasn’t enthused by LLMs) must move out for Apple’s AI…and even its own consumer release software…to improve.
What I don't understand is how I can, allegedly (I've never tried), ask siri to dig through my photos and iMessages to look for stuff, but when I ask it to summarize the website on the screen it says it can only see literally what's on the screen and can't scroll down or grab the entire article that's already been loaded.
Those limitations come from a "protecting the user" mentality. That they know better what you should be allowed to do than you.
If Apple forces developers to support Shortcuts, ensures true Mac-to-iPhone parity, and lets users generate scripts using natural language AI, it’s game over. Nailing this native integration would single-handedly kill the market for almost every derivative AI app out there.
Siri-generated Shortcuts would be so good if they were good.
Interesting notes about Craig's management style in that article. I have worked with a few C-Level folks with similar traits (e.g. "smartest in the room", etc.) and never seen great outcomes. Have been hoping for years that MacOS would get back to being a delightful upgrade -- faster, better, etc. For me, that has been lost during the Craig decade or so. Note: This is just an opinion, and not well informed apart from seeing a guy like guys I worked with. I.e. not Craig's fault, and it might have been worse if not for him.
I have an iPhone 16 which supposedly was made for Apple Intelligence. I enabled Siri intelligence beta and all requests slowed to a halt, even calling someone. I’m guessing that everything was being processed in the cloud as opposed to my device. It couldn’t even detect that I had a timer running that it started.
So I disabled it and rely on normal Siri again because somehow it’s better.
It's a crowded field and Apple has almost no expertise in AI, but hopefully they can find some angle that makes use of their massive locked-in customer base and huge piles of cash.
I'm not sure about "no" expertise in AI. A lot of their AI work has been in efficiency and getting models running on the fanless rectangles everyone has in their pockets. Same for Google. They just don't seem to be interested in developing big, expensive frontier models. They've leaked CLAUDE.md in software releases a lot, so for internal development it seems they're fine paying Claude API prices while Anthropic burns piles of cash on training and advancement.
They don't need frontier models. They can just take what Muse is and make Siri that.
Or just make Siri useful at all.
What Apple needs to accomplish with AI, given their lack of a home-grown frontier model, hasn’t been done, at scale and with refinement, by any company anywhere, except maybe Perplexity. They’re struggling to refine a unified consumer AI platform that is built as an abstraction layer.
I'm now using Grokbot and a custom written equivalent regularly in my day to day, in fact I'm so enamoured with this way of working that I'm actively building it into most of my new projects and tools... meanwhile I can't get Siri to respond to me reliably any more. This whole space is moving so fast a yearly update cycle isn't going to be enough.
That said, I could see there being a free Siri, and a paid for version that is far more powerful. Add it into Apple One? It's also worth noting here that I'm over caring whether this runs on device, I care more than it works.
Rather than making posts/threats about how they're going to ramp up their permissions systems even further... instead they should be scrambling to turn a conversation with your phone a first class interface, otherwise people will discover tools that live outside the Apple ecosystem and then people will gradually move away from it (I can't see them actually opening it up unfortunately). This will quickly become an existential threat to them, because as great as the iPhone is.. if it doesn't fully work with a user's AI agent(s) it will be a problem for Apple.
apple needs a better software cadence- siri cannot wait for every WWDC to get a bump, we should be seeing it every month in this current AI field
As much as a legend this guy is, it's the call of the CEO to "retire" him. He would have pursued a more ambitious ai strategy if he really wanted to, so just urging him to do that again won't cut it.
Apple pushing top talent to competitors reminds me of Apple 1994.
They're in an interesting spot with AI, any rival has to run behind Siri to do the same things but they're not guaranteed a 30% piece of their revenue anymore, they're leaning on Google for datacenters so there's a real cost for usage but they're not yet worth paying to use.
> better integrated with its technology than the competition
And even this is looking more like short-term relief than a lasting advantage, and in a lot of countries they've already exhausted that option.
https://archive.is/MjIRk
I think we've long ago lost the plot on software. I great this news with continued apathy.
Software and ceiling-less capitalism are at odds with each other.
I would like at least the basic things to work. Like spelling/auto-correct on my iPhone. You'd think after 10+ years of using the same apple account across many devices, it could learn to spell the word I've used 100 times already? It looks worse than the simplest markov chain. Not to mention finding files in Finder - impossible. I tend to instead use find/grep from the terminal, and worst case, when I can't remember exactly the name (but have some rough ideas of what it may be called), I get Codex to find it for me.
The subtext is clearly that Federighi may need to reconsider his position.
Not before time. Apple's software is a bland directionless nothingscape compared to the hardware.
An even minimally usable AI product would be a good start…
I don’t get why journalists keep calling appointed bureaucrats “czars”.
That absolute power fetish begun in second Trump term and didn’t seem to age well.
Do they expect any of those czars to kill any of their children to protect their position?
Fair enough, some of US czars actually had their private firing squads. And they did kill people at their will. And remained unpunished.
But for the love of the people I wouldn’t bow to people like Bovino in a tech article.
Anyways, I can’t imagine that’s what US people want.
It dates to the 1930s under Roosevelt and became common again in the 1970s and has remained fairly common since then:
https://en.wikipedia.org/wiki/Czar_(political_term)#Modern_u...
https://en.wikipedia.org/wiki/List_of_U.S._executive_branch_...
Affordable DRAM forecast just moved to 2050 ...
It is quite difficult for me to think of a mass-consumer AI business here. You can add a hundred little features scattered throughout iOS that give real improvements, and maybe put that behind a subscription?
For platforms like instagram, you can use it to drive way more ad revenue and it sorts itself out, as described here: https://x.com/firesidealpha/status/2107082123215810862
What is the equivalent for iOS/iCloud?
fm serve is a good start.
> Another concern is Federighi’s preference for planning over relatively short periods, typically 12 to 18 months, compared with the multiyear road maps favored by Apple’s hardware organization.
I feel like everything I read about Apple's AI strategy is full of these kind of "huh?" moments. How can someone who wants Apple to pursue an "Ambitious AI Strategy" simultaneously believe that the software org can produce meaningful multi-year roadmaps right now?
The future of software and technology was already looking pretty bleak pre-AI. Now the AI slopification of everything feels like the final death blow to me.
certainly changes the landscape for professional work (for the worse mostly), but I'm having a blast in my personal work.
Really? AI is finding and fixing critical bugs that have existed for decades. Hobbyists are recreating and remixing closed source games that have been abandoned. It has never before easier for a total amateur or an experienced coder to create as capable tools as now.
It is very disappointing what Apple has done in the past two years with Siri and AI.
First they did an ad for Apple Intelligence and Siri. Then two years silence. Then they launched it, it’s worse than it was before, not available in the EU and they use models from other companies.
Paywall
archive.ph
[dead]
And?
Apple’s opportunity is the greatest hope I know of for the democratization of AI. Fingers crossed
Democratization as in buying a narcistic president with a golde iphone, amongst others?
Why would a $5 trillion dollar company have any interest in democratizing anything at all?
Boggles my mind that people are fooled into into thinking a company has their best interests at heart.
They're a hardware company that made trillions of dollars selling hardware. "Commoditize your Compliment" is one of the oldest business strategies out there.
their share of income from services and subscriptions has been increasing steadily year over year, i think they're interested in keeping that growth
What do you think the word democratize means, exactly? It's a serious question.
Here's what I think it does not mean: walled gardens, vendor lock in, closed source, subversion of interoperability.
A hardware company whose strongest gross margins come from services and has been trending in that direction for over a decade. Heck, if trends continue, their margins and profits might be predominantly services based in the not to distant future.
And so the enshittification of Apple begins.