Phib comments on Foom seems unlikely in the current LLM training paradigm

[Phib]

Hi, writing this while on the go but just throwing it out there, this seems to be Sam Altman’s intent with OpenAI in pursuing fast timelines with slow takeoffs.