FAIR really does stand out from the field on available weights, safety considerations that seem at least plausibly connected to abuse potential, and a long-term

FAIR really does stand out from the field on available weights, safety considerations that seem at least plausibly connected to abuse potential, and a long-term, serious academic research agenda. And I’ll remind you that their ocean of multi-modal, longitudinal over decades training set is not in any “uncertain” state as concerns copyright. I’m pretty fricken annoyed with e.g. Boz given the fact that people at least sort of believe Carmack and Palmer now, which means they at least sort of believe me now. But that increases, not decreases my obligation to be fair, and Meta AI is on fire.