> the industry is only starting to catch onto on-device models language models. i mean there are technical limitations and tradeoffs to running LLM-size models

> the industry is only starting to catch onto on-device models language models. i mean there are technical limitations and tradeoffs to running LLM-size models locally. doesnt help to ascribe it to lack of foresight when it is a known Hard Problem.