On: Apple Intelligence for iPhone, iPad, and Mac

[drexlspivey]

That’s a necessary temporary step until these powerful LLMs are able to run locally. I’m sure Apple would be delighted to offload everything on device if possible and not spend their own money on compute.