Apple announces 'Apple Intelligence':personal AI models across iPhone, iPad, Mac (opens in new tab)

(9to5mac.com)

8 pointsanujbans1y ago13 comments

13 comments

I think the most important and somewhat massive news is that Apple built an entire LLM private cloud using Apple Silicon - that's a big deal/investment.

talldayo1y ago

Almost makes you wonder if they're inferencing models on that hardware but not training it there. They worded that part of the presentation very carefully.

Which also begs the question; how much money would Apple save by using Nvidia for everything? Probably not much since they don't have to pay margins on Apple Silicon bought for themselves. But I suspect there is a literal monetary cost to bruteforcing an Nvidia-scale server network with weaker hardware.

buildbot1y ago

Yeah I’d guess inference, training relies much more on high speed interconnect between nodes, which I’m sure they could do but it’s certainly another step up in complexity.

dmix1y ago

I noticed Tim said "LLM and diffusion models", is diffusion models just the image generation stuff?

buildbot1y ago

Yeah diffusion models are typically for image gen

cube22221y ago

This seems really cool.

They said the models can scale to "private cloud compute" based on Apple Silicon which will be ensured by your device to run "publicly verifiable software" in order to guarantee no misuse of your data.

I wonder if their server-side code will be open-source? That'd be positively surprising. Curious to see how this evolves.

Anyway, overall looks really really cool. If it works as marketed, then it will be an easy "shut up and take my money".

cromka1y ago

It's still a matter of how secure their hardware will be. So there's no warranty they or, rather, the authorities won't snoop.

gigel821y ago

They need a lot more information about how "Apple Intelligence" actually works. They can't just wave the magic wand and say "some" of the functionality runs in the "private" cloud - we need one single checkbox to say we want the cloud or not and strong guarantees things don't accidentally get cloud-ified.

OtherShrezzing1y ago

They've named this platform "Personal Intelligence Service"... Or PIS. Which seems like the type of brand name you'd find in a GTA game.

cromka1y ago

Publicly verifiable server workers to run larger models? It's interesting how private it will actually turn out to be.

ls6121y ago

Sounds like some combination of source-available server software and remote attestation from the Apple servers to me.

cromka1y ago

It's still a matter of how secure their hardware will be. And there's no warranty they or, rather, the authorities won't snoop.

jcfrei1y ago

Is Apple running a proprietary LLM or are they licensing one from OpenAI, Google, etc?

1 more reply

j / k navigate · click thread line to collapse

13 comments

buildbot1y ago

I think the most important and somewhat massive news is that Apple built an entire LLM private cloud using Apple Silicon - that's a big deal/investment.

talldayo1y ago

Almost makes you wonder if they're inferencing models on that hardware but not training it there. They worded that part of the presentation very carefully.

buildbot1y ago

Yeah I’d guess inference, training relies much more on high speed interconnect between nodes, which I’m sure they could do but it’s certainly another step up in complexity.

dmix1y ago

I noticed Tim said "LLM and diffusion models", is diffusion models just the image generation stuff?

buildbot1y ago

Yeah diffusion models are typically for image gen

cube22221y ago

This seems really cool.

I wonder if their server-side code will be open-source? That'd be positively surprising. Curious to see how this evolves.

Anyway, overall looks really really cool. If it works as marketed, then it will be an easy "shut up and take my money".

cromka1y ago

It's still a matter of how secure their hardware will be. So there's no warranty they or, rather, the authorities won't snoop.

gigel821y ago

OtherShrezzing1y ago

They've named this platform "Personal Intelligence Service"... Or PIS. Which seems like the type of brand name you'd find in a GTA game.

cromka1y ago

Publicly verifiable server workers to run larger models? It's interesting how private it will actually turn out to be.

ls6121y ago

Sounds like some combination of source-available server software and remote attestation from the Apple servers to me.

cromka1y ago

It's still a matter of how secure their hardware will be. And there's no warranty they or, rather, the authorities won't snoop.

jcfrei1y ago

Is Apple running a proprietary LLM or are they licensing one from OpenAI, Google, etc?

1 more reply

j / k navigate · click thread line to collapse