[coffeefirst]

Agreed.

The ideal implementation of AI for Apple is probably to finally make Siri work. This isn’t necessary fancy, just let me set some calendar events without knowing the magic words or tell it to open Overcast and play the new Gastropod episode. Better yet, for power users, let me set up reusable shortcuts using natural language.

The most important part of this is it doesn’t necessarily feel like AI. The user does not like AI for its own sake or the weirdos who ramble about putting them into a permanent underclass. The user likes messaging their friends and playing music.

To much of this hype cycle has no user in mind.

[samrus]

Absolutely agreed. It feels like tech companies forgot that they are supposed to add value to users. Theyve been shoving random AI usecases down their users throats with no regard for whether it works for the users flow or not. When theres so much value to be had from AI in normal products. Claude code is the best in this right now, probably because the engineers themselves are users.

This isnt unprecedented, its what happened in the dotcom bubble as well. But then that tech started getting used properly as well. So i think its a matter of time before claude code levels of value is avialable to normal users

[the_snooze]

>Absolutely agreed. It feels like tech companies forgot that they are supposed to add value to users.

They lost the plot long ago. They're firmly in extraction mode now: how much value can they get from end-users?

[fuzzfactor]

Like many new things, or newly marketed things, the natural acceptance factor is lower than the proponents would need in order to fulfill their dreams.

Beyond that though is the dream that highly persuasive efforts will be effective at overcoming hesitation and converting it into new desires and preferences. Like the way it has worked under so many situations. But with survivor bias firmly in mind, those are the orgs where no miracle was actually required before it could lead to a windfall.

[new_account_100]

> When theres so much value to be had from AI in normal products.

Please elaborate

[mrob]

Fuzzy search

Reverse dictionary

Stack Overflow clone, except you're guaranteed to get an unreliable answer promptly instead of waiting for a human to give it

OCR, with new and exciting failure modes

Machine translation, with new and exciting failure modes

Endless possibilities for exploiting the stupid and ignorant while destroying the web in the process

Note that only the first two are unalloyed good, and they can be done with embeddings without generative AI.

[MrDarcy]

Replace search for one.

[kjkjadksj]

Proper search is far more powerful. I can set the weights for the results to various parameters, focusing on certain metadata in particular. Sad to see popular search tools have gone stupid in recent years. But search is still very powerful and Imo ai is no replacement for good search. An example of a powerful search engine is pubmed and the logic you can craft in your queries.

[CSSer]

Right. It's deterministic, and determinism should be the goal. It's not metaphysical. Some users know what they want while others do not. The software we create (by any means) should give users who know what they want the tools to find it, and guide those who don't until they do. Software exists to help us create our fate. It surprises me how many people are willing to relinquish that control or never wanted it, even within our ranks, by using AI to simplify experiences. IMHO, the optimization for most, but not perhaps not all, tools is to introduce AI internally to refine, create and expose more parameters, not less. Search is a perfect example of this.

[marcus_holmes]

Search is great when you know what you're looking for.

I want a "how do I?" function alongside search that will explain their product to me. Especially since so many SaaS products have absolutely terrible UX - it looks lovely, but you cannot discover anything, and you cannot intuit how to do something. Menus auto-hide, scroll bars don't work so you don't realise there's another half of the page you're looking at, buttons don't have tooltips or any explanation of what they do, icons are lovely but don't actually describe the thing they do, colours are lovely but I'm colourblind so aren't helpful, there's no useful help page for "this is how to do the really obvious thing you're trying to do...", or at least not one that I can find using the search terms that make sense to me.

I think I'm at 1 success out of 15 attempts for Gemini to explain how to do something in Google Sheets/Docs, though, so I'm not hopeful that anyone can actually implement this.

[sincerely]

>I think I'm at 1 success out of 15 attempts for Gemini to explain how to do something in Google Sheets/Docs, though, so I'm not hopeful that anyone can actually implement this.

People love to talk about this as one of the helpful features of AI (knowledge extraction from documents/summarizing), but I'm really not convinced. The last generation of models seem to have 70-90% accuracy on tasks like this, which is way below what i'd consider a reliable tool

e.g. https://www.frontiersin.org/journals/digital-health/articles...

https://pmc.ncbi.nlm.nih.gov/articles/PMC11197181/

I don't know if there are any benchmarks for this sort of task, maybe the new ones are improved but I also doubt that people are using GPT5.5 pro ultrathink for these tasks anyways

[what]

Except the LLM runs a bunch of searches and summarizes what it finds… what happens when they have no search engine to piggy back off of?

[tobr]

> This isn’t necessary fancy, just let me set some calendar events without knowing the magic words or tell it to open Overcast and play the new Gastropod episode. Better yet, for power users, let me set up reusable shortcuts using natural language.

Isn’t this the proverbial ”faster horse”? Ie let me do exactly what I can do now, in a very slightly different, possibly very slightly more convenient way?

[kibwen]

If the user asks for a faster horse and you sell them a car, you win.

If the user asks for a faster horse and you sell them a trebuchet, you lose, no matter how fast the trebuchet would technically get them to their destination.

[paulhebert]

Also, some people just want a faster horse

[fuzzfactor]

Yes, don't forget if the user asks for a faster horse and you sell them a faster horse you still win.

And you may be able to sell them what they are already asking for a lot faster than what they are not.

Now if you are trying to sell them something that they would rather not even have at all, that's another story too.

[BobbyTables2]

Others really want the trebuchet…

[simonh]

That’s fine if the vendor anctually understands the users needs better than they do, otherwise it’s just pure chance if they happen to align.

[layer8]

A car is still a faster horse in the sense that you decide where to go with it. It doesn’t outsource decisions to somewhere else.

(Arguably the car affords you better control than an unruly horse. Self-driving cars are moving us closer to the horse again. ;))

[fuzzfactor]

>Self-driving cars are moving us closer to the horse again.

Maybe there is some parallel to the way that AI is moving "cutting edge" programming closer to the mainframe/dumb-terminal paradigm.

[joseda-hg]

A good horse is quite the autopilot

I've seen a horse get a black out drunk rider "safely" to a hammock and then continue to it's stable to rest

[basisword]

Nothing wrong with a faster horse when AI isn't reliable enough yet to produce the car. Don't leave me with my aging horse while you cross your fingers that something better might come along someday.

[noashavit]

I don't think we're at an aging horse moment, but a hurt one. One that can run fast sometimes but limps to the finish other times

[flashman]

No, it's asking for my two-legged horse to have four legs

[whatshisface]

The whole point of AI is that if something different happens, it's not you doing it.

[paul7986]

Exactly as a UX person who watched the movie H.E.R. A few times i feel something like is the next UX internet evolution… talking and texting to AI where all visuals need to be seen appear your iPhones lock screen. Siri would in the background communicate with AI agents of businesses to govt organizations to organizations to your friends and family to get things done for you. Lessen the need to unlock your phone and Apples creating AI AirPods just use the iPhones lock screen to create/show the appropriate visuals and text.

As UX / UI professional of 17 years I think design is a dying field the above would kill digital UI design quicker. Yet the UX would be less steps / friction to complete tasks which is the harbinger of UX design…less is more.

On a side note I’m just in medical school studying a mid level Concentration. I don’t foresee a LONG term future in digital design and development much anymore.

[kjkjadksj]

Now I’m picturing a dystopia movie. No one knows how to plan any events anymore, some event appears in their calendar and they show up and find some people there matched to their profile. People get silenced from certain events and can’t get back in. Like a personalized music playlist but it’s your entire life. People forget how to organize and create original ideas, and any prospect of revolution becomes as likely as expecting a farmers cattle to rise up.

[paul7986]

Could be where things go…UX is all about less is more ..less steps. Time will tell.

[paulhebert]

That’s the simplistic view of UX thats commonly sold.

Less steps isn’t always better. Friction has its place.

A basic example is an “Are you sure?” confirmation before a destructive action.

I wish there wasn’t so much focus on “less clicks.” It’s often to the company’s benefit at the detriment of the user

[fuzzfactor]

Not only that, but to many users the whole point of a computer is that it can do the same exact thing every time.

[basisword]

Exactly this. I use Siri for two things: remind me x at date/time and set a timer for x. And it even screws these up 10% of the time. If you make those work flawlessly but it also works with any app on the device I'm sold. I'd even buy a new device if it was limited to that. Let OpenAI and Anthropic worry about changing how we work in a revolutionary way. Whatever the outcome is there people still need great products to do ordinary things and that's where Apple has always excelled.

It's crazy to me that even with a strong accent ChatGPT can nail my voice messages. If Siri can suddenly do that (and there's no reason it can't anymore) the device becomes much more useful to everyone that doesn't speak English and doesn't have an American accent.

[m3kw9]

They keep banging on Siri hoping for a different outcome is insanity by definition. Voice is actually not a very good UI for most things, it isn't very private, it's prone to mistakenly think I'm talking to it, and is bad for dense info/info organization. Siri should only be activated very very deliberatly, not "Hey siri", and don't make it act like jarvis because you will not in the near future with the smarts it needs.

[janalsncm]

> Voice is actually not a very good UI for most things

Agreed. But it is a good UI for some things, and which things is probably situation and user specific. Many people’s frustration with Siri is that many of the things it should be good at based on their decision to try, Siri cannot do.

[ikeke]

Voice is only good in the context of something that is high priority(e.g talking to a customer service agent) / highly satisfactory (e.g talking to a friend in real time).

Otherwise humans hate that interface.

[macintux]

Setting aside the blind and others with challenges for whom voice has the potential to be (if it isn’t already) a magical interface, I find it great for my HomePod for very narrow use cases. Wake me at 7am. Tell me the weather forecast for the weekend. Have I received any new messages?

[janalsncm]

Voice is a good interface for people who can’t or shouldn’t look at a screen. CarPlay for example, could have a lot more powerful voice capabilities.

[nonethewiser]

>The ideal implementation of AI for Apple is probably to finally make Siri work.

Upon learning about LLM's however many years ago (3? 4?), literally my first thought was:

"Oh, how Siri is supposed to work."

It's the single most obvious application.

[altmanaltman]

When that happens, it will notify the end of the hype cycle. Knowing about tech and working in it, you might have a very different perspective on AI and its limitations, but the companies are still literally selling it as if it's some magic black box that'll solve every problem and make humanity 100x more efficient. If we reduce that to a "smarter Siri that makes fewer errors", the public perception will be massively disappointed. They have promised AGI and people have bought that promise given their valuation, walking back on it = suicide.

[adastra22]

That's what AGI is though..

[zitterbewegung]

People keep on saying Apple is far behind its competitors on AI. If Apple just waited on their Apple Intelligence announcements about Siri or other features that would have been best. Right now Apple makes money off of any subscriptions through the App Store which is actually profitable compared to the foundational AI companies which are spending trillions to make a technology which everyone will have but no one will expect to pay the cost of making the technology.

[tomaskafka]

Yep. I am pleased to say that after over a decade of development and high tens of billions USD in costs, Siri can finally play Knights of Cydonia in Youtube Music at a first try. "That Queen song from Sonic hedgehog movie" will probably need another 20-50 billion $ and I am happy to let Ternus spend it.

[kjkjadksj]

Do people want that? I mean I don’t think it can discern if I say 15 or 50. Why would I leave that to chance that the ai properly grokked my message when despite what I’m guessing decades of work in the speech to text field, it is still pretty unreliable? Doing the task myself is trivial enough and 100% reliable.

[airstrike]

In an ideal world in which LLMs behave as advertised, the idea is that you'd be delegating to the agent what you might otherwise be doing yourself. And just like when delegating to an intern who called you collect from a payphone, you may need to spell out "one-five" when you say 15 to make sure they don't hear 50.

[matty22]

If they could get "Hey Siri, play 'Song Title' by 'Band Name'" to open the Music app and play the song I ask for, that's the extent that I give a damn about AI being introduced to iOS.

[xerox13ster]

The only thing I have ever wanted from a voice assistant or AI assistant or anything like that on the phone is the ability to be listening to music, interrupt the music and tell it “add ‘X by Z’ song next in now playing” and then have it add it and continue playing the same playlist, but it’s never done that. It always replaces the entire now playing list.

[JumpCrisscross]

> ideal implementation of AI for Apple is probably to finally make Siri work

Wouldn't the simplest solution be to auction off Siri's back end the way Apple does Safari's search bar in iOS?

[benoau]

They're reportedly already doing that, AI services will be able to publish "Extensions" that Siri will use and then those services can compete amongst themselves to power it.

But this is contingent on the same services not being able to replace Siri and being able to reserve its APIs for Apple's exclusive use, and they have a pretty tenuous grasp on that these days.

https://9to5mac.com/2026/05/05/ios-27-will-let-you-choose-be...

[robotresearcher]

Apple has already announced that they’ll be using Google Gemini.

https://blog.google/company-news/inside-google/company-annou...

[danaris]

No, because what Siri does needs to be tightly integrated in ways that search does not.

[inadequatespace]

>Absolutely agreed. It feels like tech companies forgot that they are supposed to add value to users.

They literally have forgotten as they’re just following orders from investors (public or private).

Before those orders were “get as many users as possible” as valuation is based on credibility of reaching TAM.

Now the orders are “just use AI as this will be necessary to stay relevant and avoid being displaced” in addition to of course “profitability now”

[wolvoleo]

Or at least let me give multiple commands.

"Hey Siri turn on the livingroom lights" "Hey Siri set the thermostat to 19'

Being able to go "Hey Siri turn on the livingroom lights and set the temperature to 19" would be so much easier.

For a real AI this would be no issue. But Siri is completely hand scripted.

[bonesss]

I have a grander vision for an ideal Apple “AI”: anti-AI.

I’m picturing a combination of on-board facilities and online services from the Apple cloud that Apple product holders could use to flag and filter LLM slop. As a value added prospect, iPhone users who read HN or used TikTok would be seeing clear UI-level indications of when they’re interacting with slop with options to kill it.

In my estimation it would provide platform benefits without losing capabilities, leverage Apples hardware and not advertising positioning, fix critical issues of spam and scams, and let them market a higher calibre of online experience. Also, they could un-eff Siri - “play album X starting at track Y”, come on, it’s 2026.

[WillAdams]

The thing which kills me is a lot of this was working back in the Newton days.

[smugma]

Can you expand on this? Having used two different Newton models, even squinting, I don’t understand what you’re getting at.

[WillAdams]

On my Newton MessagePad, one could write things such as "Lunch <name> Friday" interact with it using a stylus to activate "Newton Intelligence" and it would create a calendar event for the next Friday, and attach the contact as a link.

[Rover222]

I JUST got a notification from Google Home that Gemini is now available for Hey Google stuff. Finally.

[jcgrillo]

Earlier today Siri notified me over and over again to message a particular contact on the GMail app. I have no earthly idea why this was important enough to notify me about, I can figure out whether I need to message people. It provided no hints what the contents of the message were supposed to be, or why I might need to message them now instead of some other time. 20yr ago when I worked in an industrial steel fabrication shop, a couple times a week someone would exclaim "shoot the fuckin' engineer that came up with that one!" usually regarding some bone-headed physically impossible weld on a plan or a procedure that would clearly result in an assembly being out of tolerance, but sometimes something more serious like a weak or dangerous design. Now we're well into the "shoot the fuckin' engineer" stage with tech products. Abusing people like this is wrong. Their attention is a finite resource. Their reward centers are vulnerable. Mindlessly pushing the buttons in the psychological control room of vast, diverse, and increasingly stressed populations is a profoundly stupid idea. If we're not careful somebody is actually going to start shooting the engineers. Would they be wrong to? I don't think the backlash against this shit is going to be small or subtle, and I'm honestly afraid to be associated with this industry right now. Y'all are playing with fire in the most reckless, clueless, thoughtless, and callous fashion. Be better. Stat.

[wpm]

Shortcuts already lets you set up custom workflows you can trigger by voice.

[coffeefirst]

I know. I actually use it a lot. But the way you assemble shortcuts leaves a lot to be desired for anything nontrivial.

[protocolture]

I felt the same way about NFT's. Its a cool protocol. Theres some cool stuff that could hypothetically be built on the protocol. Selling people NFT's felt like trying to sell someone a TCP or a DNS. The protocol is not the product lmao.