H

Habryka

CEO @ Lightcone Infrastructure
21405 karmaJoined Working (6-15 years)

Bio

Head of Lightcone Infrastructure. Wrote the forum software that the EA Forum is based on. Often helping the EA Forum with various issues with the forum. If something is broken on the site, it's a good chance it's my fault (Sorry!).

Comments
1362

Topic contributions
1

It would be great if you could provide evidence (beyond your word) for that! Even saying that you talked to people at OP, or any other epistemic status would be helpful.

I have talked to multiple people at OP and close to OP who seem to agree that OP is very hesitant to fund anything right-coded. The correlations are extremely obvious, Dustin has made relatively concrete statements to this affect, and I really can't reconcile this kind of extremely sparse and confident public communication with the very obvious and clear feedback I get from people working closely with OP and looking at OPs actual granting track record. 

I am also frustrated with then Max giving as a counter-example a Norwegian think tank, which of course has nothing to do with what I meant by left/right coded, since what it means to be right or left coded of course is drastically different in different countries, and the underlying cause here is US political polarization and reputation management, which does not generally extend to foreign countries.

Look, it's really hard to provide any kind of commentary or transparency on organizations like OP. The communication around the whole Dustin/GV/OP shift has been extremely limited, and the power-dynamics are extremely tense and messy. It really doesn't help to have someone show up and just plain contradict something I said without any further evidence, arguing purely from authority. 

Like, what is the next step of this conversation supposed to be? I have shared my observations, I commented extensively on why I believe what I believe, and I clarified what I mean by my statements in a huge amount of detail, only for you to show up and give a contextless "This is false". I think it's useful for you to share what you believe, but I think it's really clear that in this domain it is extremely rarely appropriate to just make a blanket statement like this. At least say something like "I don't currently think this is true" as opposed to this weirdly aggressive, authoritative and contextless statement from high up.

Ideally you would say something like "while it is true that OP has become much more hesitant to fund right-leaning political organizations, I think saying that OP does not want to fund 'anyone even close to right of center' is too strong. It is true there is a large left-leaning bias, but I think we will see OP overcome those in many cases if something looks good enough by theirs and Dustin's values, such that describing it as much as a hard line as you are doing here seems more heat than light-producing". 

Like, I am pretty sure you believe something like this, because you are not blind and you see the same evidence as I have, but your comment sure does not communicate that.

What you linked to is a Norwegian Think Tank. 

Yes, it has "right of center" in the text of the article you linked, but of course my commentary was about US politics, and a Norwegian think tank doesn't interface with that. What is "right of center" in Norway is completely different from what is "right of center" in the US.


Commenting on the broader topic brought up by the top-level comment, I sent over the spreadsheet of all grants from Open Philanthropy in 2024 to GPT-o1-preview asking the following question: 

Here is a spreadsheet of all of Open Philanthropy's grants since January 2024. Could you identify whether any of them might meaningfully constitute a grant to a "right of center" political think tank or organization within the United States?

GPT-o1-preview responded with: 

After reviewing the list of grants provided, none appear to meaningfully constitute a grant to a "right of center" political think tank or organization within the United States. The grants are primarily directed toward universities, research institutions, and non-profit organizations focused on areas such as global health, animal welfare, artificial intelligence safety, and effective altruism. While some organizations are U.S.-based, none are identifiable as right-of-center political think tanks or organizations.

It's not perfect (and I just used the first wording that came to mind, which might be off in various ways), and of course it's been less than a year since there was a large shift in how Open Phil related to GV and as such is under tighter constraints here, and so we shouldn't expect there to necessarily be much counterevidence even if the underlying observation is false, but I don't think there currently exist any counterexamples, or at least none that are publicly available.[1] 

  1. ^

    (Asking for confirmation, GPT-o1-preview said: 

    Based on the list of grants provided, there do not appear to be any organizations or grants that would serve as a counterexample to the assertion:

    > Open Phil does not want to fund anything that is even slightly right of center in any policy work.

    All the grants listed are directed toward organizations that are generally considered non-partisan, centrist, or left-of-center in the United States. They primarily include universities, research institutions, and non-profit organizations focused on areas such as global health, animal welfare, biosecurity, artificial intelligence safety, and effective altruism.

    No grants are identifiable as supporting U.S. right-of-center political think tanks or organizations. Therefore, based on the information provided, the assertion appears to hold true within the context of U.S.-based organizations.

    )

I don't think the benefits would outweigh the enormous costs, no. I think there is space in EA for election discussion, and indeed things like the Personal Blogposts are a decent fit for that, as are other spaces that are higher trust (like sessions at EA Global). It's not like this topic is banned, its just disincentivized, which seems very reasonable to me.

The personal blogpost category is pretty clear. On hover it says:

There are very few topics that are as difficult to discuss rationally as US partisan politics. It very blatantly obviously is the kind of topic that tends to destroy the sanity of large swaths of otherwise smart and well-reasoned people. What is a topic that would be more deserving of "difficult to discuss rationally"? 

I am not sure whether I agree with the categorization here, but I don't think there is any hypocrisy or inconsistency in the EA Forum in making this decision.

I think this is the closest that I currently have (in-general, "sharing all of my OP related critiques" would easily be a 2-3 book sized project, so I don't think it's feasible, but I try to share what I think whenever it seems particularly pertinent): 

https://www.lesswrong.com/posts/wn5jTrtKkhspshA4c/michaeldickens-s-shortform?commentId=zoBMvdMAwpjTEY4st 

I also have some old memos I wrote for the 2023 Coordination Forum I would still be happy to share with people if they DM me that I referenced a few times in past discussions.

CEA seems to maintain control over most high-level aspects of EAGx, so I don't think this counts as competition.

The answer for a long time has been that it's very hard to drive any change without buy-in from Open Philanthropy. Most organizations in the space are directly dependent on their funding, and even beyond that, they have staff on the boards of CEA and other EA leadership organizations, giving them hard power beyond just funding. Lincoln might be on the EV board, but ultimately what EV and CEA do is directly contingent on OP approval.

OP however has been very uninterested in any kind of reform or structural changes, does not currently have any staff participate in discussion with stakeholders in the EA community beyond a very small group of people, and is majorly limited in what it can say publicly due to managing tricky PR and reputation issues with their primary funder Dustin and their involvement in AI policy.

It is not surprising to me that Lincoln would also feel unclear on how to drive leadership, given this really quite deep gridlock that things have ended up in, with OP having practically filled the complete power vacuum of leadership in EA, but without any interest in actually leading.

Oh, that's an interesting idea. In general seems good to take newsletters and digests and summaries and use other platforms for additional reach.

Agree! I am a bit worried the discussion is a bit social-drama heavy on LW, since it's kind of the only outlet for that kind of stuff on the frontpage, but that in itself is also somewhat of a success (having a place for more community-oriented discussion without taking over the whole site).

  • GiveWell, which takes a combined broad and HNW direct fundraising approach, seems to have hit some limiting factors in 2022 after having grown rapidly for more than 10 years.
  • Similarly, growth of The Life You Can Save, Effektiv Spenden, Animal Charity Evaluators, and Giving What We Can (all largely broad direct fundraising organisations at the time) seems to have stagnated somewhat at around the same time, suggesting this may have had something to do with external factors (e.g. the economic downturn and/or the FTX crisis), but there could also be other factors at play here, e.g. target groups becoming saturated.

I will take bets at relatively high odds that these external factors were the reason for the reduction in growth. Approximately anything EA-adjacent stopped growing during that period.

Load more