Sooo....how to prevent abuse with privacy model? I mean someone can change porn actor face with their "friend" right?
Last edited:
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Gurman: Apple Working on On-Device LLM for Generative AI Features
- Thread starter MacRumors
- Start date
- Sort by reaction score
Client-side [device-resident] processing does seem to be the key 'must-have' for the M4 products.
The I/O with the data on the Internet another key requirement. But if they get that right, then I've found reason to upgrade, although I might wait to see the user feedback from the early adopters.
The I/O with the data on the Internet another key requirement. But if they get that right, then I've found reason to upgrade, although I might wait to see the user feedback from the early adopters.
Hopefully they have a new set of developers working on this. Keep anyone associated with Siri far far away.
I'm really excited for iOS 18. Considering Siri currently needs internet access even to play one of your downloaded playlists on Apple Music, this should be quite the leap.
According to this rumor, their AI's LLMs are on device which is safer than any cloud AI.
Whoever wrote that must be so ignorant if they don’t understand that an LLM like Gemini running on device would make the device unusable, hot and dead in 30 minutes. You can only have a custom cut down LLM made for smart phones if you want fully on device.
When Apple eventually gets up an running, I would think that they will more that likely follow suit and offer subscription based services similar to the ones that ChatGPT provides: Free, Plus and Team. The choice of subscription type could potentially mean the introduction of imposing limits to the utilization of on-device AI capabilites across all Apple devices.
Wanting to monetize yet another digital service is not only likely. It is inevitable. Perhaps they will call it "AppleAI".
Wanting to monetize yet another digital service is not only likely. It is inevitable. Perhaps they will call it "AppleAI".
HobeSoundDarryl
macrumors G5
So higher RAM for this AI to use or rolling out some kind of AI RAM in iCloud with a forever rent component? If "onboard" means on board the computer or iDevice, doesn't RAM have to be big enough for it?
Someone tell me why I should want or need generative AI?
#OldManYellsatCloud
#OldManYellsatCloud
Google Search didn’t have that effect, so I doubt it.
If they weren’t behind, they’d be behind like the others who’ve poisoned their LLMs through training via Reddit and Twitter. They are learning from others who were too quick to market, I hope, and building something from the ground up predicated on Apple values. Temporarily using others is a necessary evil.
The Siri developers left Apple long ago, which is why it is in the state it’s in.
The problem with AI from OpenAI, Google and Microsoft is that they all used the same logic with their LLM to train them. It was all based on doing it fast and getting it out there because of the competition. What they all did with their training is to be more concerned with quantity instead of quality. It is the basic GIGO Problem.
I don’t know, and I’m not saying Apple will be better, but it appears with Apple there is at least a ray of hope
I don’t know, and I’m not saying Apple will be better, but it appears with Apple there is at least a ray of hope
Last edited:
I am not expert on AI, but how on earth will mobile devices represent the multitudes of free parameters needed to make AI useful? Something tells me the phone will pre-process data and then phone home to a computer cluster for final AI output.
Apple itself is no saint. Bending over to Xi China to do business there. Apple always puts profits first. Every else is marketing
What I want is for Siri to be able to use the apps and information local to the phone that it already has access to in an intelligent way.
Right now I've been trained to not even try Siri. For several years they've claimed more on-device functionality but it's still functionally useless for anything but doing things like toggling settings.
They released a paper on how to shrink subsets to be used on-device. If it's pre-trained in the cloud on the data types present on the iPhone, especially easy in this case since the iPhone silos information according to Photos, Reminders, Calendars, etc. it should be possible to use that on-device for basic functionality.
They could make a smarter pre-trained Siri that could actually do useful things.
But I'm not getting my hopes up.
Whatever it is one thing is for sure…it will be A bsolutely I ncredible.
Apple has a massive task to pull this off. Siri is terrible, it was terrible before chatGPT, despite the way they have talked about it. Apple are going to have to be really really Apple about their move on this; there’s basically no wiggle room.
I'm naive when it comes to neural engines and AI in general but my hope for M4 and onwards is that apps that are introducing AI - Adobe, for example - can take advantage and perform complex tasks blazingly fast. Also that FCP, etc, are optimised with new features. It'll take time, I know, but if the M4 comes out and in house apps like FCP haven't been optimised then that's a misfire.
Don't forget that with Apple Silicon your iPhone is faster and more powerful than most Windows Desktops!
It doesn’t make sense to train it on user data as such. Rather, it makes sense for it to have selective access to user data, just like Siri has.
Given that a lot of consumers don't want a in your face AI interface. It doesn't take a genius to assume that consumers just want something that is usable not that it pertains to AI empowering how you use a computer. This is a lot like demanding quality of user experiences not that you work faster.