I polled of 105 people at EAG on Friday evening, asking them the question: "Which field are you in, or more importantly want to be in?"

I estimate I polled around a third of the people in the King George 3 room in The Brewery between 18:30 and 19:30 on the evening of Friday 15th April.

I categorized people's answers into some arbitrary categories. Here are the raw results:


Or, counted up and put into a more readable format:

32 — AI Alignment / AI governance

15 — Community building

11 — Don't know / nothing yet

10 — Animal welfare / advocacy

10 — General longtermism / bio risk / s-risk

7 — Law or policy

5 — Global health / longevity / mental health

5 — Founding charities / movements / companies

4 — Climate change

4 — Earning to give

1 — Improving institutional decision-making

1 — Operations ("whatever is needed")

So about 30% of people in the room were in AI.




Sorted by Click to highlight new comments since:

At that time there were only two events, and only for part of the time: a parents meetup (18-19) and a high-school student meetup (19-20). So no simple to adjust bias I could see :)

Not sure if you are in fact seeing this, but presently I see 3 posts with a similar title. The two previous ones had "105" in the title. Just making sure you know this. Also, thank you for posting this. Quick survey results are usually nice to see. 

I didn't know that thanks - have removed those. I clicked the submit button and saw nothing happen

Are there more comprehensive surveys of what cause areas EAs are working in? Rethink's EA survey just seems to ask the cause areas EAs think are most important.

I don't know, and I'm definitely interested if you find out

Another obvious way to do this I guess is count on SwapCard. Any thoughts on the differences in output between the two techniques?

I guess you’re selecting for “shoe leather” or something for being there physically.

Also, I’m not sure what event this is, but some people prefer events and some prefer meetings. It would be interesting if that preference is systematically different between cause areas. But I expect this to be subtle I guess. Even if this was an adequate sample it would be hard to infer too much.

Yeah I'm not claiming anything much - it was just for fun 😊

Curated and popular this week
Relevant opportunities