TLDR

• CEA’s University Groups Team worked on a wide range of projects this summer^[1]
• We wrapped up a round of UGAP and OSP (our programs supporting university group organizers with digital resources, mentorship, training, and for UGAP a stipend) and launched a new one. Both rounds had over 100 people participating across both programs
• We ran a summit for relatively experienced group organizers, had four interns join us (who worked on projects like updating the Groups Resource Centre and setting up residencies), and continued to give out (relatively small) group support grants, to cover operational costs for university groups
• We’re excited about potential future projects, like supporting more ‘top’ universities and/or seeding more support for AIS groups, but continue to reflect on what projects would be most valuable

Intro

CEA’s University Groups Team is working on supporting university groups focused on EA ideas, and their organizers. We wrote more about our mission and vision last year here - note it is slightly outdated. Our team currently consists of Jessica McCurdy (team lead), Jake McKinnon, and Joris Pijpers^[2].

In this post we share what we’ve been working on this summer, including launching new rounds of UGAP & OSP, a summit, an internship program, and more. We won’t touch much upon what we did in spring ‘23^[3] (but have a retrospective of UGAP fall ‘22 here), and have also chosen to not expand in detail on why we are focusing on these programs and not others, in favor of getting the post out sooner. Additionally, we are still in the process of deeper impact analysis^[4]. We hope to publish more details on these in the future.

Overall, we are pretty happy with what we were able to achieve this summer and are excited about future directions.

Why university groups?

We think university groups have the potential to be especially promising places to introduce people to EA ideas, and then help them learn more about and act on them:

• University group members (and organizers in particular) have a strong track record of making significant contributions to EA priorities, both in and beyond meta work
• Students are usually in the process of thinking through their priorities for their careers and lives. In this process, they are often open to exploring new ideas and communities, and have time to do so
• Universities are places where people build communities and social networks. These in-person social ties are important for people being more likely to take significant action on EA ideas
• Universities can have high concentrations of pre-screened talent
• Universities are a very large source of new EAs each year, which means that student groups might have an outsized influence on the development of EA culture and movement priorities. Setting norms of good epistemics through advice and content aspirationally may improve that of EA

A note on which universities we focus on

We think top universities, such as those previously listed as focus universities, have a high concentration of top talent. That means that we extend higher touch support to top universities and have some programming focused on them specifically (like our summit and residencies). However, the majority of talented, altruistic and greatly influential people do not come from top universities. Therefore we think scalable support (like UGAP & OSP) which includes both top and non-top universities is also worthwhile. We have many examples of people who are doing work that we think is really important, who did not come from top universities, and might have not gotten involved with EA ideas if it wasn’t for their university group. We also think people that we help from these universities tend to be particularly counterfactual.

Separately, we should note that EA is a rapidly changing ecosystem. Changes to bottlenecks in key priorities and cultural norms within EA influence our programming and focus. We are regularly reflecting on what the current landscape means for our programming. This means we might seem to be changing our focus and programming more than one might expect^[5].

What we've been working on

UGAP & OSP

UGAP

The University Group Accelerator Program (UGAP) is a remote program that helps to make starting new EA university groups easy and accessible. We provide students with a mentor, digital resources (e.g. an outreach guide), training, and sometimes a stipend. Since the start of the program in 2021, we have supported over 200 group organizers, coming from every populated continent. This round, we have organizers participating in UGAP from schools including IIT Bombay, University College Dublin, University of Florida, University of Sao Paulo, University of Warwick, and many more.

OSP

The Organizer Support Program (OSP) is a mentorship program aimed at university EA group organizers to help them prepare for the start of the semester. Similarly to UGAP, it provides student group organizers with a mentor, resources and training, but no stipend. We ran a very low touch pilot of OSP in the ‘22 spring semester, as we noticed pre-existing EA university groups can also benefit from resources and mentorship - especially if there’s a lot of new organizers running the group. We have now also accepted a few AI safety groups, to explore if our programs are able to add value for those groups. This round, we have organizers participating in OSP from schools including Stanford University, Pontifical Catholic University of Chile, Harvard University, Cambridge, Delft University of Technology, the University of Hong Kong, and many more.

Want to be notified when new rounds of UGAP or OSP open? Sign up here!

UGAP & OSP spring ‘23

This summer, we wrapped up the spring ‘23 rounds of UGAP and OSP, and made a number of improvements to both programs. You can find a short retrospective, which includes an overview of what improvements we made, in Appendix 1. We share a breakdown of participants by country in Appendix 2.

UGAP & OSP fall ‘23

This summer, we launched the fall ‘23 rounds of UGAP and OSP:

• 147 applications: 85 for UGAP and 62 for OSP
• 111 participants^[6]: 43 in UGAP (with 36 receiving a stipend) and 68 in OSP^[7]

Again, we share a breakdown of participants by country in Appendix 2.

A note on our bar for admission

Due to the decrease in funding in the space, our bar for organizers that we support has gone up. This means that some organizers, who in an ideal world we would like to support, do not receive support from us. Application processes are not perfect, and sometimes we make mistakes in both directions. We are planning on reevaluating our bar this winter based on how correlated our evaluations of applications are with organizer outcomes. Our selection criteria are similar to that of the Open Philanthropy University Organizer Fellowship, though since our support is less money intensive and more introductory, we usually have a slightly lower bar than them.

What’s next for UGAP & OSP?

• We’re considering switching to needs-based grants for UGAP (as participants can apply to Open Philanthropy’s University Organizer Fellowship), but we need to spend more time on looking at whether this is practically feasible
• We’re considering merging UGAP and OSP, to simplify the application process and reduce confusion
• We are considering building out a track for AI safety university group organizers

Group organizer summit

In May, ahead of EAG London, we hosted a group organizer summit in collaboration with CEA’s Partner Events Team (thank you to michel for all his work on this!). The retreat aimed to bring together (mostly) experienced university group organizers from established EA university groups, such as those previously listed as focus universities. More specifically, the goals of the summit were to facilitate lasting coordination between organizers at established groups, help organizers improve their group’s strategy by learning from guests and their peers, and to kick-off collaborative projects and spark more enthusiasm for EA uni group organizing.

Results

• The guest LTR^[8] (likelihood to recommend) was 8.43/10, but the attendee LTR was 7.68/10
• Attendees reported a mean of 7.13 new connections
• The median counterfactual use of time attendees reported was 3.5x

Despite a relatively low LTR compared to similar events, our best guess is that this event created important coordination and relationships that otherwise wouldn’t have happened, like a workshop for AIS organizers, more coordination on residencies, and more communication between group organizers. However, we think there are various things we can do to make future versions of similar events go better. Appendix 3 includes some potential improvements for future similar events.

Summer internship

Our team hosted a summer internship. We received over 100 applications, and ended up making 7 total offers. Five people accepted an offer (four of which ended up working with our team, and one with the Virtual Programs (VP) team), one person accepted another opportunity but will contract with us part-time for the foreseeable future, and the other also chose another opportunity but now works with us as a UGAP/OSP mentor. We also passed on promising applications to other teams at CEA, which led to one more person interning with the VP team.

What our interns focused on:

• Sam Robinson has been updating the EA Groups Resource Centre - the new website is now live!
• Luise Wöhlke improved the back-end of our UGAP and OSP application processes
• Lauren Mangla and Peter Gebauer coordinated residencies, where one or more experienced EA community builders travel to a university at the start of a semester to help the EA or AIS student groups get off to a good start or learn from that group
• Irene Hulsen was interning with CEA’s Virtual Programs Team, where she helped with piloting a Career Planning Program

Groups support funding

As part of CEA’s group support funding, the CEA University Groups Team gives out grants for expenses of EA university groups.

• Between August 2022 and June 2023 (the 2022-2023 academic year for the northern hemisphere) we made 120 grants to 91 groups, totalling $329,809 USD
• So far, for the 2023-2024 academic year (July 1st - September 22), we made 24 grants to 24 groups, totalling $47,537 USD

Other things we did

• We attended various EAG(x) conferences and retreats, and also ran sessions there
• We are in some coordination with regional and national EA organizers, and help them with ‘seeding’ university groups
• We publish various (public and non-public) memos, like “Where on the continuum of pure EA to pure AIS should you be?”
• We have had a few hundred chats with uni group organizers (though much of this is through our mentorship programs)

What we didn’t run

We consciously made the decision to run quite a lot of programs this summer, which meant that we did not have a lot of slack time to jump on potential new opportunities, focus on improving our object-level understanding of problems in the world, or engage in more communication with group organizers. In hindsight, we think this was roughly the correct decision for this summer, but going forward we want to improve our responsiveness if we were to again opt in to having a high predetermined workload.

With the help of various stakeholders and group organizers, we brainstormed a lot of potential projects. Of course, we unfortunately don’t have the capacity to run them all, but here^[9] are some things we would be interested to see in the space by us or by others.

What’s next

We’re currently figuring out what our priorities should be for the upcoming ~9 months. We think that the following are among the areas we are most likely to focus our efforts on:

• Wrapping up and evaluating the fall rounds of UGAP & OSP, and likely launching a new round for spring ‘24
  • We are considering running a workshop/retreat for some of the fall ‘23 UGAP & OSP participants (and potentially others)
• Expanding our support for the traditional highly ranked universities
• Further exploring what role our team could (and/or should) play in the AI Safety university group space, which might include building out programs within our team or extending support to others in the space^[10]
• Exploring new potentially valuable projects; we expect to spend at least 0.5 FTE piloting a new project

We’d be excited to hear your ideas on what we should focus on next. In general, if you have any questions, ideas, or feedback, reach out to us on unigroups@centreforeffectivealtruism.org! Thank you for your continuous support :)

Appendices

Appendix 1: UGAP & OSP spring ‘23 retrospective

Key results

This summer, we wrapped up the (northern hemisphere) spring ‘23 rounds of UGAP and OSP. Some key stats:

• 130 applications: 54 for UGAP and 76 for OSP
• 103 participants: 33 in UGAP and 70 in OSP
• LTR of 8.94 for UGAP (n=19) and 7.29 for OSP (n=34)^[11]

Improvements to UGAP and OSP after spring ‘23

To improve our programs, we’re always gathering feedback from participants, mentors and stakeholders. Based on feedback on our spring ‘23 round, we made the following edits:

• Improved the quality of mentor matches, by better taking into account time zones and asking mentors and mentees to reflect on their match (and indicate if they want a switch) after their first three meetings. Some early evidence suggests that the matches are better^[12]:
  • The UGAP participants’ overall satisfaction with their mentor went up to 9.61 for the fall ‘23 round (according to the 3rd week (~mid-program) survey, n=38), compared to the 9.09 rating for the spring ‘23 round (according to the mid-semester survey, n=23)
  • The OSP participants’ overall satisfaction with their mentor was 9.44 for the current fall ‘23 round (3rd week survey, n=43), versus 8.74 for the spring ‘23 round (mid-semester survey, n=35)
• Luise Wöhlke helped us drastically improve our backend systems during her internship with us, which streamlined our admissions process, improved our communication, and allows us to better track participants in our programs
• We made improvements to the UGAP resource guide, designed an OSP guide, and improved the kickoff sessions for both programs
  • Overall LTRs went up: from 9.00 for the UGAP spring ‘23 round (mid-semester survey) to 9.34 for the UGAP fall ‘23 round (3rd week survey), and from 7.97 for the OSP spring ‘23 round to 8.84 for the OSP fall ‘23 round
• We drafted improved templates for recurring mentor meetings and a semester planning sheet. We also improved resources for mentors, including the mentor training

Appendix 2: UGAP & OSP spring ‘23 and fall ‘23 participants - breakdown by country

Appendix 3: Potential improvements for future group organizer summits

• Increase transparency and better manage applicant expectations in our marketing. For this round we weren’t clear enough about the target audience for this event, which caused some disappointment and confusion with applicants and stakeholders
• Provide more structure to participant-led sessions, to help them address macro strategy questions. We initially thought we would not have much to offer as we had not been involved with more experienced organizers for over a year, but we now think we could have provided more structure
• Keep the first day of the event structured with non-participant-led sessions to allow attendees to ease into the event
• Improve session voting prior to the event. (Concretely, use Slido instead of Slack to propose sessions before the event)
• Have a better memo session proposal workshop that gives participants a chance to take a stance on a key issue and think about sessions they could lead
• Ensure participants and guests mingle more, especially earlier in the retreat
• Be even more proactive about checking in with participants leading events to ensure they have a good plan
• Being proactive about communicating with participants when we think they will mostly be providing value to others rather than getting value themselves

1. ^{^}

   Summer in the northern hemisphere, that is

2. ^{^}

    We are part of CEA’s Groups team, and also work with contractors who help us

3. ^{^}

    Note that spring in the northern hemisphere is fall in the southern hemisphere, and vice versa. For brevity, we’ll refer to the northern hemisphere seasons/semesters in this post

4. ^{^}

    As a sneak peak, we are collecting information on significant actions such as whether someone has been interviewed for an EA organization, worked full time at an EA organization, gotten 80k career advising, taken the GWWC pledge and more. Early data suggests that the GWWC pledges and 80k calls taken by even a small portion of respondents justify the costs of our programs alone, but there are many tricky counterfactuals we are working with here. GWWC pledges and 80k are also not the main things we are tracking

5. ^{^}

    For instance, we are aware of the increasing interest in people able to fill more senior positions. We still think targeting university students is valuable, but could be convinced to decrease our focus there

6. ^{^}

    Unfortunately, we don’t accept all applicants to our programs. We sometimes reject participants because they hand in an incomplete application, or because we think community building might not be their comparative advantage, but most applicants get rejected because we believe they are not sufficiently able or comfortable at talking about complicated EA ideas yet and should probably learn more first

7. ^{^}

    Our OSP acceptance rate is higher than that for UGAP for a few reasons: we marketed OSP less publicly, OSP has no stipend, OSP mentorship can be capped at 3 weeks (which means it takes less resources), and people applying to OSP are generally more experienced

8. ^{^}

    We don’t think LTR is the best way to measure value or impact, but we find it useful in comparing programs over time and measuring user satisfaction

9. ^{^}

    Note this comes from an internal brainstorm, so the document is not very polished

10. ^{^}

     We are aware of branding considerations around CEA supporting AIS groups, and are keeping this in mind as we explore ways to improve this space

11. ^{^}

     We also asked about overall satisfaction, with average ratings being a 7.92 for UGAP and 7.29 for OSP

12. ^{^}

     There are two problems with comparing these numbers: firstly, we asked participants for a likelihood to recommend their mentor in the spring ‘23 survey, but for their overall satisfaction with their mentor in the fall ‘23 survey. Secondly, the 3rd week survey for the fall ‘23 round came earlier in the semester than the mid-semester survey for the spring ‘23 round