I would not go so far to say simply having proximity to OpenAI is problematic as long as the instructor manages to convey that their decision to associate with a leading AI company is controversial (or at least the more general point, that many reasonable people find what the AI companies do to be highly irresponsible). I somewhat trust that this is handled reasonably in this case. Overall, I think the worry here is less that people who might only be somewhat safety-pilled give courses like this, and more that people who are very safety pilled don’t have the opportunity to do so (due to selection effects in CS/ML academia).
I would not go so far to say simply having proximity to OpenAI is problematic as long as the instructor manages to convey that their decision to associate with a leading AI company is controversial (or at least the more general point, that many reasonable people find what the AI companies do to be highly irresponsible). I somewhat trust that this is handled reasonably in this case.
Overall, I think the worry here is less that people who might only be somewhat safety-pilled give courses like this, and more that people who are very safety pilled don’t have the opportunity to do so (due to selection effects in CS/ML academia).