Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Siri AI

apple.com

Post reply on HN
Print view View original

441–450 of 746 posts

Re: Siri AI

#441

Post by Barbing »

meatmanek wrote: post #213
bilsbie wrote: post #53

It’s weird it says I can ask Siri about a document in front of me but can I ask it about a webpage I’m currently reading? (It’s been driving me crazy there’s no “AI this” button to discuss whatever is on my screen.)

I'm guessing they'll integrate with the double-tap-the-bottom-of-the-screen feature that pulls up siri in front of a screenshot. Currently it doesn't seem to hook into "visual intelligence", and needs to call out to ChatGPT to do anything with the screen contents.

Tangential note on

> double-tap-the-bottom-of-the-screen feature that pulls up siri

It’s disabled if not using Apple Intelligence, and can’t tap screen while talking to Siri (it dismisses instead).

Now they’re gating features to the M3 I’m not convinced wouldn’t work on expensive Apple Silicon predecessors… am more convinced the double tap disable is intentional.

Re: Siri AI

#442

Post by idle_zealot »

zmmmmm wrote: post #438

It's interesting how anemic the use cases seem to be - we see the same things recycled over and over: "reword my email", "remove object from picture", "add a reminder", "summarise my text message which was already only 20 words long" etc etc. As if these are the major problems in people's lives. I really feel like there's a fascinating valley of death between simple things that actually work and things of real value…

Even the aspirational use cases you're talking about basically are just "digital secretary." There's a massive problem with that even if the models end up being capable in the future. The value of a secretary is that you know them, they know you, and you trust them to do things right. There are stakes if they don't. No company can provide that as a service at scale for everyone without it being a disaster. Not because it's not technically possible, but because of the incentives. That much power over the details of so many people's lives is irresistible; there will be persistent temptation to use it. The presence of that possibility makes the secretary impossible to trust.

Re: Siri AI

#443

Post by kennywinker »

nozzlegear wrote: post #427
kennywinker wrote: post #418

Earlier quoted context omitted.

Using AI to summarize a message by saying what the message says in the same amount of words as the message So helpful!

Is your complaint that the example text message in the image wasn't three paragraphs long? It's just showing search results.

My complaint is this image is supposed to tell the story of how this stuff is useful or new. Finding a message from earlier in the same day and then paraphrasing the message instead of just showing it isn’t useful or new. A simple keyword search uses infinitely fewer GPUs and finds me the info as fast without the risk of hallucinations or datacenter driven environmental collapse.

But for whatever reason - because this stuff is mostly useless, because they’ve all got ai psychosis, or because of general dysfunction in the corporate structure - they can’t come up with a better example.

Re: Siri AI

#444

Post by cootsnuck »

zmmmmm wrote: post #438

It's interesting how anemic the use cases seem to be - we see the same things recycled over and over: "reword my email", "remove object from picture", "add a reminder", "summarise my text message which was already only 20 words long" etc etc. As if these are the major problems in people's lives. I really feel like there's a fascinating valley of death between simple things that actually work and things of real value…

Yup, precisely. Turns out getting AI to be reliable at doing useful things is harder than we've all been led to believe by the dominant narratives.

https://www.normaltech.ai/p/new-paper-towards-a-science-of-a...

Re: Siri AI

#445

Post by Barbing »

manwe150 wrote: post #293
onesociety2022 wrote: post #277

Earlier quoted context omitted.

I never understood how any regulatory body is going to decide which APIs in iOS must be made available to third-parties to hook into. So what if I'm a third-party maker of TCP/IP stack and I want Apple to offer me the ability to sell my custom TCP/IP stack to my iOS customers as a replacement for the stock TCP/IP stack that ships with iOS. Clearly no regulatory body has cared about that because it's too niche of a sp…

Replacement TCP/IP stack sounds like a VPN—which iOS allows

True. Will add, device must be supervised to use VPN always-on which is possibly sensible albeit annoying (would have to reinstall iOS and set up as new I believe).

Re: Siri AI

#446

Post by KawaiiCyborg »

Danox wrote: post #230
KawaiiCyborg wrote: post #131

Earlier quoted context omitted.

As a consumer it also doesn't seem right that Apple can just use all their private APIs that no other company is allowed to use to tell me what I can and can't use on my phone. If I want Anthropic to have the same level of access that Apple grants itself on _my_ phone, I should be able to do so.

Don’t buy it you have Android which is 3/4 of the world? What’s the problem? You also have Windows which is also 3/4 of the world.

As if Android and Windows don't also use the same tactics. Windows used to not let you uninstall Edge and Android used to just default to Chrome and Google Search, now they have to prompt you what you actually want to use. If the big companies are left unchecked, of course they'll use their size to bully other players to increase their own profit, but always at the cost of the customer.

Re: Siri AI

#447

Post by KawaiiCyborg »

aurareturn wrote: post #226
KawaiiCyborg wrote: post #131

Earlier quoted context omitted.

As a consumer it also doesn't seem right that Apple can just use all their private APIs that no other company is allowed to use to tell me what I can and can't use on my phone. If I want Anthropic to have the same level of access that Apple grants itself on _my_ phone, I should be able to do so.

If I want Anthropic to have the same level of access that Apple grants itself on _my_ phone, I should be able to do so. I want to install my personal software on my refrigerator, washing machine, hand shaver, coffee maker. I bought them. They're mine. /s If people want to buy open hardware, then just buy those. If they don't exist, make them yourself.

Do you rather want your refrigerator, washing machine and coffee maker send all the data it collects about you during your usage to some company you don't even know? How is it ridiculous to want to be able to make your devices do what you want to do, not what the big company behind it wants?

Re: Siri AI

#448

Post by kennywinker »

usef- wrote: post #428
kennywinker wrote: post #419

Earlier quoted context omitted.

> It could be 200 texts back. Ok? But if I search for something just show me the thing I searched for (the proof). There is literally no need to repeat it in a slightly different tense. Who is that helping?

If I asked you over voice what Mike said he wanted for his birthday, it's usually less confusing if you just answer the question ("he wants snow boots") rather than started reading out words from another person ("I love it here at the snow, ...")

If that’s the story you’re trying to tell, where is the query in the image? The image alone doesn’t tell that story.

But also:

a) do you trust the llm to get it right 100% of the time? Because i’m gonna always read the original message to make sure.

And b) just excerpt the message “mike says i love it here at the snow … but i need snowboots” if you’re so desperate to shoehorn LLMs into everything, that’s just as easy a task for them as summarizing is.

Re: Siri AI

#450

Post by cjonas »

idle_zealot wrote: post #442
zmmmmm wrote: post #438

It's interesting how anemic the use cases seem to be - we see the same things recycled over and over: "reword my email", "remove object from picture", "add a reminder", "summarise my text message which was already only 20 words long" etc etc. As if these are the major problems in people's lives. I really feel like there's a fascinating valley of death between simple things that actually work and things of real value…

Even the aspirational use cases you're talking about basically are just "digital secretary." There's a massive problem with that even if the models end up being capable in the future. The value of a secretary is that you know them, they know you, and you trust them to do things right. There are stakes if they don't. No company can provide that as a service at scale for everyone without it being a disaster. Not becaus…

These use cases will just be built as "open source" (openclawd) or even custom one off application in the future. I've been building apps to run the tedious parts of my life recently. Meal planning, personal finance, bills, tax organization... Why would I pay for services that will be enshiftified when I can build a app that does exactly what I want in an afternoon. Yes the code is shit and it wouldn't scale... But it doesn't need to
Post reply on HN