My current thoughts on the technical AI safety pipeline (outside academia)
- There is room for something like RAISE
- AI safety technical pipeline does not teach how to start having novel thoughts
- AI safety is not a community
- Mixed messaging regarding independent thinking
- There is pressure to rush into a technical agenda
- Mass shift to technical AI safety research is suspicious
- AI safety lacks a space to ask stupid or ballsy questions
- AI safety is harder than most things
- Nobody understands what makes people snap into AI safety
- Giving advice in response to generic questions is difficult but important
- Newcomers in AI safety are silent about their struggles
- It is difficult to find people to bounce ideas off of
- It is difficult to get feedback on published work
- Ongoing friendship and collaboration is important
Appendix:
