In my regular research (behind a paywall), I have been saying for a while that I think the future of AI is not large language models (LLM), but small language models (SLM) run on local desktop computers or even mobile phones.
a ton less people actually use a desktop computer that would allow use for an slm. the world has moved on from desktops, sorry. those same users, however are sure as fuck already users of LLMs whether they want to know it or not. its being piped in as tooling fucking everywhere.
you arent doing that kind of mass scaling using slms. ‘4 out of 5’ use cases is not going to be running on your local desktop, nope. thats just ignoring the practical implimentations that exist now let alone in 5 years.
i watched broadcast tv for the first time in a long, long time and its amaing how 25% of ads seem like they are for ai driven services. no desktop required.
this feels shortsighted.
a ton less people actually use a desktop computer that would allow use for an slm. the world has moved on from desktops, sorry. those same users, however are sure as fuck already users of LLMs whether they want to know it or not. its being piped in as tooling fucking everywhere.
you arent doing that kind of mass scaling using slms. ‘4 out of 5’ use cases is not going to be running on your local desktop, nope. thats just ignoring the practical implimentations that exist now let alone in 5 years.
i watched broadcast tv for the first time in a long, long time and its amaing how 25% of ads seem like they are for ai driven services. no desktop required.