Quite likely paving the way for that, and on the software side models continue to get more efficient as well. I think it’s pretty realistic to expect that you’d be able to run equivalent of full blown models that need expensive hardware today on a home computer within a couple of years.
Wow!! Is this the beginning of widespread local AIs?
Quite likely paving the way for that, and on the software side models continue to get more efficient as well. I think it’s pretty realistic to expect that you’d be able to run equivalent of full blown models that need expensive hardware today on a home computer within a couple of years.