I’d imagine that the labs growing in size will diminish the importance of AI safety people within them, though. Look at all the new people joining Anthropic/OpenAI recently—most are highly accomplished but don’t have long track records of caring about AI safety. In the case of OpenAI there have also been a lot of departures of the original core of employees who were initially motivated by EA/AI safety concerns (some left due to these concerns ofc.)
I’d imagine that the labs growing in size will diminish the importance of AI safety people within them, though. Look at all the new people joining Anthropic/OpenAI recently—most are highly accomplished but don’t have long track records of caring about AI safety. In the case of OpenAI there have also been a lot of departures of the original core of employees who were initially motivated by EA/AI safety concerns (some left due to these concerns ofc.)