4 ms·
Have you considered that LLMs in their current form just…aren’t very good or useful? That may be why Apple isn’t using them yet.
by callalex 3y ago
Have you considered that LLMs in their current form just…aren’t very good or useful? That may be why Apple isn’t using them yet.
- balder1991 3y agoThis is my guess too. They’re using it on their iOS 17 keyboard for better text prediction though.
- appplication 3y agoDespite all the hype, and being genuinely intriguing, this is the truth. They can do some neat tricks, but overall utility is low. Model hallucination in particular is a problem. I don’t see Apple rolling out any model that has the potential to lie to you.