Besides Ilya Sutskever, is there any person not related to the EA community who quit or was fired from OpenAI for safety concerns?
Gretchen Krueger quit recently: https://x.com/GretchenMarina/status/1793403475260551517
She has a 2-year-old EA forum account https://forum.effectivealtruism.org/users/gretchen-krueger-1 , has written reports in 2020 and in 2021 where most (all?) co-authors are in the EA community, is mentioned in this post, and is Facebook friends with at least one CEA employee
tbf it’s pretty hard to do any work in AIS without coauthoring things with EAs at least sometimes, for better or worse (probably worse).
It’s possible to work at OpenAI and care about safety without being friends with CEA staff though.
It doesn’t seem that anyone OpenAI besides the EA community is too worried, which to me is a positive update.
Potentially Pavel Izmailov– not sure if he is related to the EA community and not sure the exact details of why he was fired.
https://www.maginative.com/article/openai-fires-two-researchers-for-alleged-leaking/
Some other people like Andrej Karpathy and Ryan Lowe have left in the same time period, but have avoided use of safety-based justifications, and so far as one can tell it’s unlikely that safety was the reason there.
@Zvi has a blog post about all the safety folks leaving OpenAI. It’s not a great picture.
They all seem related to the EA community, and for many it’s not clear if they left or were fired.
Besides Ilya Sutskever, is there any person not related to the EA community who quit or was fired from OpenAI for safety concerns?
Gretchen Krueger quit recently: https://x.com/GretchenMarina/status/1793403475260551517
She has a 2-year-old EA forum account https://forum.effectivealtruism.org/users/gretchen-krueger-1 , has written reports in 2020 and in 2021 where most (all?) co-authors are in the EA community, is mentioned in this post, and is Facebook friends with at least one CEA employee
tbf it’s pretty hard to do any work in AIS without coauthoring things with EAs at least sometimes, for better or worse (probably worse).
It’s possible to work at OpenAI and care about safety without being friends with CEA staff though.
It doesn’t seem that anyone OpenAI besides the EA community is too worried, which to me is a positive update.
Potentially Pavel Izmailov– not sure if he is related to the EA community and not sure the exact details of why he was fired.
https://www.maginative.com/article/openai-fires-two-researchers-for-alleged-leaking/
Some other people like Andrej Karpathy and Ryan Lowe have left in the same time period, but have avoided use of safety-based justifications, and so far as one can tell it’s unlikely that safety was the reason there.
@Zvi has a blog post about all the safety folks leaving OpenAI. It’s not a great picture.
They all seem related to the EA community, and for many it’s not clear if they left or were fired.