this kind of story is compelling to believe
but this only makes sense if certain labs have SOTA capabilities that differentiate them - and are enduring
the observed reality is that any advantage a lab develops - is quickly caught up to
๋ ๋ณด๊ธฐ
many folks seem to believe the full extent of the ai race is publicly observable
for better or worse, at least 3-4 ai labs (excluding openai, anthropic and deepmind) have sota capabilities they may never share externally
completely economically rational
๋ ๋ณด๊ธฐ
