In December, we ran a “Next Steps” retreat for promising intro fellows. This post summarizes what we did, why we ran it, what went well, and what could be improved in future retreats. 

I think other student group organizers should consider running retreats after intro fellowships. 

If you are interested in running one, this Notion page includes several valuable resources (e.g., our schedule, COVID protocol, post-retreat survey, etc.). You may also want to watch this video by Diana Lim (one of the retreat attendees).

What was the “Next Steps” retreat?

The Next Steps retreat was a 2-day retreat for promising intro fellows from Bryn Mawr, Haverford, Swarthmore, and UPenn. The retreat consisted of discussions and talks led by student group organizers and EA professionals.

Harry Taussig (Haverford) and Olivia Jimenez (Columbia) were the primary organizers. I (UPenn) also helped.

What were the goals of the “Next Steps” retreat?

Broadly, the primary goal of the retreat was to help intro fellows feel motivated and prepared to work on the world’s most pressing problems. More specifically, we hoped that the retreat would help intro fellows:

  • Form connections with EA professionals and other intro fellows.
  • Learn more about "EA careers" (i.e., by meeting some EA professionals, learning about their paths, and understanding what they do).
  • Reflect on key uncertainties as they consider their own careers.
  • Identify actions they want to take after the intro fellowship.
  • Feel more ambitious and motivated.
  • Learn more about the EA community, EA landscape, and EA orgs.

What went well?

This section is partly based on my own impressions and partly based on information collected in the attendee feedback form. I would want to see these repeated in future retreats:

  • Inviting EA professionals to give talks about their work (exposing intro fellows to people who are doing exciting/impressive direct work seems extremely motivating, and I don’t think they get a strong sense of this during the fellowship. It also makes “EA careers” seem less weird/unconventional and more realistic/attainable).
  • Paired one-on-ones (especially pairing intro fellows with EA professionals or with organizers from different schools). Attendees frequently mentioned the value of forming connections with other EAs.
  • Running on time. We generally started and ended things exactly on time.
  • Talks and exercises that described strategies to combat fear, friction, imposter syndrome, etc. (see slides of a talk from the retreat: “Turning Beliefs into Actions”. See also resolve cycles, goal factoring, and fear-setting).
  • Providing information about the “current state of EA” (e.g., funding situation, major EA orgs, major EA hubs, EA jargon, examples of ambitious projects that are currently being launched).
  • The schedule was strong. In general, the talks were relevant and well-prepared.

What could have been better?

I think a few things about the retreat could have been better, and I would want to see these changes in future retreats:

  • The invite list could have been more selective. I think this would have led to higher-quality interactions and a tone that was more ambitious/serious. At the same time, I think this is pretty hard to get right, because you want to avoid false negatives (i.e., you don’t want to exclude people who actually would have taken the ideas seriously). Still, I think for this particular retreat, we could have been more selective about the attendees. My guess is that about 40-50% of the intro fellows were highly engaged & taking things seriously. Getting that number closer to 80 or 90% would have considerably improved the retreat.
  • Ideally, we would have had EA professionals that covered a wider variety of topic areas. We had 4 professionals; 2 were pursuing policy careers and 2 were pursuing movement-building careers. The retreat would have been better if we had a wider array of EA professionals (e.g., an AI safety professional, a biosecurity professional, a global priorities researcher, and a policy person). Note that I think this is pretty tough, especially if the retreat is not happening in a major EA hub. Aim for the ideal, but it’s fine to just get whoever is around/willing to attend.
  • We could have done a better job planning unstructured social activities during the evenings. One evening, we played “hot seat,” which felt a bit exclusive (i.e., it led to conversations that involved a lot of inside-references that organizers knew, but it may have created a divide between organizers and fellows). I think we should have thought, in advance, about activities that would be inclusive & likely to foster strong connections between attendees. Some ideas here include board games, random nerdy stuff (discussion groups about EA topics or EA-adjacent topics), and structured bonding activities (stuff like Coat of Arms, which is a personal favorite of Harry’s!). Note that some attendees mentioned that Coat of Arms was their favorite part of the entire retreat!
  • We could have also arranged the space in a way that would have facilitated small-group conversations (e.g., set-up chairs in groups of 3-4 so that it was natural for people to have small group discussions. Rather than having one big circle of chairs).
  • On the feedback form, attendees stated that they would have preferred multiple options during the evening. Rather than having one activity as a big group, they would have preferred opportunities to get into smaller groups. This might have produced stronger connections.
  • Attendees stated that they wanted more breaks, as well as more opportunities to connect freely and casually. We could have had 10-15 minute breaks after each session. Relatedly, we could have emphasized that optional sessions are (really) optional. We tried to do this, but perhaps we could have been more explicit. Example: “You should feel free to go take a nap, go on a walk, or have a one-on-one instead of attending some of the optional sessions. Do whatever is most valuable for you.”
  • Some attendees stated that they would have preferred an all-vegan menu.
  • Some attendees stated that, even after the retreat, they still felt underqualified or not competent enough. A few stated that they felt uncertain about how they could contribute, uncertain about their futures, and concerned about career stability.

Who funded the retreat?

We received funding from the EA Infrastructure Fund.

What should I do if I want to run a retreat?

First, check out the Notion page! Second, feel free to reach out to me (akashwasil133[at]gmail[dot]com), Harry Taussig (htaussig[at]haverford[dot]edu), and Olivia Jimenez (oliviajimenez01[at]gmail[dot]com) if you have any questions.

You can also read this summary of The Art of Gathering or this Google Doc that Harry made before the retreat.

Can I see a montage video with moments from the retreat?

Yes, you can! (Credit to Diana Lim (UPenn) for making this).


I'm grateful to Harry Taussig, Olivia Jimenez, and Angelina Li for offering feedback on a draft of this post.


5 comments, sorted by Click to highlight new comments since: Today at 10:23 PM
New Comment
Some attendees stated that, even after the retreat, they still felt underqualified or not competent enough. A few stated that they felt uncertain about how they could contribute, uncertain about their futures, and concerned about career stability.

What kinds of follow-up do you think would be ideal for these attendees?

Do you have plans for follow-up on this retreat? Actually, since I'm reading this a month after it was posted; have you done any follow-up with these people already?

(Context for the question: I'm running an in person retreat and think that the question of what follow-up to do afterwards / whether we can encourage longer-term engagement is one of my biggest uncertainties.)


Btw the Coat of Arms link is giving me an "Access denied" message.

Sorry about that and thanks for pointing this out :)

Akash will update this soon!

Updated-- thanks for pointing this out, yiyang!