Disclaimers: I wrote this post in about two hours, and haven't asked anyone for feedback on it. Views my own.
I co-organised three "high-impact" career workshops for maths students at Cambridge. They were online (Nov 2020, Feb 2021, Jan 2022). They were fairly popular (100> attendees in the last one) and we also got good feedback. I think other EA groups at universities should run similar ones.
I am not a community/field builder. I am a postgraduate student in (applied) maths. I did not have any experience organising large (or small) student events before this. I was part of the student committee of EA Cambridge in the 2020-21 academic year. Maths is one of the subjects with more students at Cambridge (~250 per year). And maths / quantitative skills are very valuable in many career paths. Especially in some recommended by 80,000 Hours. So there was (and is) a large opportunity to do EA outreach to them[1].
The workshops were open to students from other subjects (we got some students from physics and computer science). We mostly targeted (and attracted) undergraduates, but we did get some postgraduates and professionals. We assumed no previous familiarity with EA. We gave them only a small introduction to EA and the 80,000 hours framework. We took this content from what at the time was EA Cambridge's general careers workshop. We tried to excite them first about a particular cause area or career path, rather than EA in general.
After the general intro to EA and 80k, we had the speaker sessions. These were roughly 15min presentations about a specific cause area / career path, together with a Q&A. They were given by external speakers working on them, who were recent maths graduates and are highly engaged with EA. The last time, we had AI Safety research, quantitative biology (for biosecurity), economics research (for GPR) and quantitative finance (for ETG). After these, we had a wrap-up section where we encouraged students to make more concrete plans based on their personal fit and career capital. In the first two editions of the workshop, we split the participants into breakout rooms for smaller discussions[2]. (If you want more details about the content, you can find them at the corresponding Facebook events https://fb.me/e/2di6G25gC, https://fb.me/e/8L6UifVtV, https://fb.me/e/3GsBzQG9Y .)
In the 2nd and 3rd editions of the workshop, we allowed attendees to jump in and out for the sessions that they were interested in. Though we strongly encouraged them to stay for the whole workshop, I think this attracted people who otherwise would not have attended.
Altogether, the workshops became very long. I think we went over four hours all three times. We had scheduled breaks, but sessions would overrun and we didn't want to stop interesting discussions. I think the lack of actual breaks was (rightly) the biggest critique we got in our feedback forms.
In the first two editions, we got about 100 sign-ups (each) and about 3o attendees (each). In the last edition (which I think we advertised better), we reached 270 sign-ups. We had over 100 participants at some point in the workshop (we had upgraded our Zoom account in preparation!) and averaged about 70 at a time. Those who filled out the feedback form (22 responses on the most recent occasion) had good comments regarding at least one of the sections. I have also personally come across some students on campus who attended and told me they liked it. Beyond that, I do not know if some of them have (or plan to) become more engaged with EA. Nonetheless, I think they were definitely worth running given the large turnup (it took a lot of time to organise it the first time, but not so much the others).
A non-negligible proportion[3] of the attendees were students at other universities. Looking forward, if we run the workshop again, it will probably make sense to do it in person. So I think there is probably potential to run similar workshops at other universities, especially those with a strong maths undergraduate program. If you might be interested in organising one of these, do get in touch.
Especial thanks to Eve McCormick for co-organising these workshops with me, they would not have happened without her. Thanks to all of our speakers: Olly Crook, Neel Nanda, Richard Ngo, Callum McDougall, Owen Cotton-Barratt, and Hazel Browne - it was amazing to have you. Thanks to Robert Harling for helping with the follow-up logistics, and to everyone who helped with advertising.
One could also make the argument that maths students are also more likely to find attractive the EA evidence-based, expected-value approach than other students.
This became unfeasible in the third edition due to the large volume of sign-ups.
I do not have the numbers, but I'd say around 40% for the most recent workshop.
