Thank you for collecting the data in these surveys! Do you have an idea for the ETA for the 2023 data?

My motivation is that when I talk to people about EA, I often hear the criticism that EA is all about AI risk, and I respond by saying that actually EAs in practice mostly prioritize and donate to global health causes. I've been linking the 2019 survey for a few years now and I'm starting to notice that it's out of date.

That makes sense! After SBF, I've wondered about the causality around charitable giving and criminality. If GP signatories are more criminal than other billionaires, that would suggest that individuals may be accepting criminality that they can justify with altruism. But if they're similarly criminal to other rich people, and rich people are more criminal than other groups, then that suggests a broader societal problem that we should adjust our legal system to properly disincentivize criminality. 

The base rate seems tricky to interpret without a real comparison. 3 out of 5 points in the summary say that the rate is high, but it doesn't seem like the evidence exists to support that claim. The implied reference class is "non EA super-rich people", but that class isn't actually examined in the data. 

The rates do seem subjectively very high. Way fewer than 10% of people I know have been convicted of financial crimes! But I wonder if founders and CEOs are being blamed for financial crimes that their companies commit, and approximately all successful companies commit financial crimes, defined broadly. 

One interesting comparison is that the percent who have ever spent a night in jail is similar to the rate for the overall American population. 

Overall great data though! It raises real concerns. 

I'm curious, do you think the Giving Pledge should have higher standards? My hot take is that we don't want to help launder reputations, but that I would personally be very happy if e.g. the Sacklers donated all their money to a real charity. 

I'm concerned that the control group doesn't seem like a representative comparison. It seems like Giving Pledge signatories are older, richer, more famous, and more successful than the sample of YC recipients. 

A better comparison group might be the world's 100 richest people, would it be possible to redo this analysis for them?

Hey @Fernando_MG, is this effort still ongoing? I've noticed that artists are paying a lot more attention to AI recently, because of the impact of generative visual AI on the industry. Are you working on ways to leverage that interest?

You should note in the intro that LMIC stands for "Low and Middle Income Countries". I love the images. 

Yeah I personally find it very hard to do ML interviews for that reason. So far I'm doing a mix of theory/conceptual questions and practical ML coding questions. It helps if the conceptual questions include some unusual setups, or ask about unusal tweaks. 

1. It depends a lot on what I'm hiring for and what the candidate's background is. If it's an ML job, I ask ML programming and design questions, but if I'm hiring someone to do networking, I'll ask a question about distributed algorithms or something. This is in contrast to how Google hires, because they're hiring generic SWEs so they don't care a lot about the particulars. 
2. If I had to hire someone more senior, I would reach out to other more senior people who I trust to ask them for help.
3. For a small team, generalism can be more important than it seems.

If there's too much demand for this, I'd also be willing to do these. 

When I was interviewing at big tech companies, it helped that I had 3 roommates who worked for Google. In general it seems like EAs who want to work for big tech but don't have existing connections to it are disadvantaged. That seems suboptimal. I'm glad you made this post.