Since the LLM stage of machine learning is quickly reaching it's maximum capabilities, what's the most promising advancement to continue advancing after LLMs stall? LCM? HRM? Some shit almost no one's heard of because we invented it yesterday?
>>106438967I would say...maybe increased storage capacity for batteries in electric cars and the like. I have no data to support that however. Just pure guess.
>>106438967I don’t think we’ll know for a while. LLMs are kind of statistical magic that honestly had no right to work as well as it does. Whatever it is though, it will have to have logic considered from the beginning.
