Comments
134
Headline finding: I audited 17 AI Safety Talent programmes. Zero of 17 have published any comparison group, rejected-applicant follow-up, matched control or randomisation. Not one. Every programme that mentions a counterfactual does it by asking participants to self-report.
Background
At least $70 million...
I think pressure towards rationalizing one's self-interest as somehow being "optimal" is not a good idea. It's better to be honest.
Singer's answer is correct. It really would be better to give more. We don't because we aren't perfect. And that's fine! Cf. My response to Caplan's Conscience Objection to Utilitarianism.