I would like to fund:

1. Talented early career AI governance folks who could be accelerated with small amounts of funding.

2. Established think tanks who would like to do work on AI safety.

In the near term, I'm excited about work on transparency measures for frontier AI developers, because I think that this will be helpful for improving the strategic awareness of governments and make it harder for an AI lab to pursue an intelligence explosion in secret.

Once there is more societal wakeup, I think there should be an international treaty to not build superintelligence. I'm excited about work on verification methods to enforce treaties, plus work on developing agreements that could become viable.