Writing this in a purely personal capacity in my effort to comment more on forum posts as I think of responses:
This is just a general meta point but, to me, this post is trying to take on wayyyy too many ideas and claims. I was really intrigued by some of them and would like to see more thorough and detailed arguments for them (ie: the fog, where are the effects, arbitrage, and the ants) . However, since this tried to make so many separate points, many claims were left unsubstantiated which decreased my confidence in the post and most single points within it. Similarly, none of the individual points felt fleshed out enough for me to engage with them here in the comments.
I am excited about creative critical critiques but generally want the caution against posting too many (unless it is framed as: here is a list of half-baked critiques, let me know which ones intrigue you and I will elaborate on them). In general, I would love to be able to point to any single claim in a post and be able to understand where that came from. However, that is not happening here. So, I'm downvoting this post but looking forward to future ones!
Hi Ivy,Just wanted to hop in re: the University Group Accelerator program. You are definitely hitting on some key points that we have been strategizing around for UGAP. I just want to clarify a few things:
Overall, CEA is planning to spend ~$1.5mil on uni group support in 2022 across ~75 campuses, which is a lot less than $1mil/campus. :)
Sorry, I was trying to get a quick response to this post and I made a stronger claim than I intended. I was trying to say that I think that EA careers are doing much more good than the ones mentioned on average and so spending money is a good bet here. I wasn’t intending to make a definitive judgment about the overall social impact of those other careers, though I know my wording suggests that. I also generally want to note that this element was a personal claim and not necessarily a CEA endorsed one.
Just another super quick response that doesn't cover everything and is purely my own thoughts and not necessarily accurate to CEA:
Just a quick response on the CEA’s groups team end.
We are processing many small grants and other forms of support for CB and we do not have the capacity to publish BOTECs on all of them.
However, I can give some brief heuristics that we use in the decision-making.
Institutions like Facebook, Mckinsey, and Goldman spend ~ $1 million per school per year at the institutions they recruit from trying to pull students into lucrative careers that probably at best have a neutral impact on the world. We would love for these students to instead focus on solving the world’s biggest and most important problems.
Based on the current amount available in EA, its projected growth, and the value of getting people working in EA careers, we currently think that spending at least as much as McKinsey does on recruiting pencils out in expected value terms over the course of a student’s career. There are other factors to consider here (i.e. double-counting some expenses) that mean we actually spend significantly less than this. However, as Thomas said - even small chances that dinners could have an effect on career changes make them seem like effective uses of money. (We do have a fair amount of evidence that dinners do in fact have positive effects on groups.)
As for your comment on funding student groups, we haven’t sent money to any group that has not asked for it. It is plausible that one of us encouraged them to ask for more since we do think it is a good use of money and would like groups to think ambitiously. We have a list of common group expenses with some tips at the bottom (including considerations on optics).
Given the current landscape, we think missing out on great people and great opportunities is a huge loss. This is especially true if you think there are heavy tails in the amount of impact individuals have. We have thought a lot about our funding guidelines, and suggestions, and feel comfortable with our current status though we are constantly reviewing and updating as the landscape changes.
We appreciate your concern and are always eager for feedback. If you (or others) want to expand on this post with a more in-depth, comprehensive version of this feedback, we’d be open to responding to this in more depth as well.
(The below is copied from a comment by Max Dalton below and I am adding it here for visibility)
"By the way, we are not planning to spend $50m on groups outreach in the near future. Our groups budget is $5.4m this year.
Also note that our focus university program is passing to Open Philanthropy."
I am not writing in an official CEA capacity but just wanted to respond with a couple quick personal thoughts that don't cover everything you mentioned
We've primarily been responding to the existing demand of group leaders running university groups, as opposed to seeding groups from scratch and we are prioritizing particularly scalable programs right now instead of bespoke support (as we wrote about in the "MVP University Group Program" in our Q3 update). There is significant existing demand for supporting new group organizers and we want to be sure to make the pathway smooth and simple for interested and prepared university groups. We expect to support the start-up stages of ~20 new university groups this semester.
Also, the Georgetown model described above relied heavily on students being able to join a virtual fellowship, which was more appealing during the pandemic when all things were virtual. However, our broad uni groups support team is evaluating opportunities to seed groups as a potential activity (and comparing the benefits per staff hour of seeding groups vs. supporting existing uni groups).
I know of one group outside of CEA that's experimenting with remotely trying to seed groups through promoting EA Virtual Programs through targeted advertising. If that's successful we'll consider incorporating it into our programs
This is very exciting! I am looking into creating and running some trainings for group organizers through CEA Scalable Uni Support :). If you or others would be interested in helping to create these, please let me know at email@example.com. I am particularly excited to hear from people who are willing to take lead on creating and running specific trainings. I think winter break is a great time for organizers to take on projects like this. For example, last winter I made the facilitator training for EAVP as a winter break project. It was a good learning opportunity for me to take ownership over a project and it turned into a helpful repeatable workshop. Sabrina's comment covers a lot of the trainings I would be excited to see. The one I am most excited to see soon is one to help train people to do 1-1s as these seem to be both particularly valuable to run and somewhat intimidating to do (at first!). Others that might be useful could be: * Movement building strategy and strategic prioritization for groups * Practicing elevator pitches for EA groups* Applied rationality for group organizers and creating positive epistemic norms
I think the big ones were the cluelessness week and the small probabilities week.
Cluelessness week pointed out that we can't really know the long-term effects of our actions. So people became suspicious that we can knowably affect the long-term future at all. This ended up being more of an empirical claim than a moral one.
The small probabilities week was challenging when put to the extreme (ie: God at your deathbed thought experiment). Additionally, some felt like the numbers of the expected future that people like Bostrom use were basically pulled out of thin air- along with the tiny probabilities of various actions affecting that future. So they were again pretty suspicious of the empirics here as well.