Thank you for posting this! I appreciated both learning how others feel about this and your thoughtful commentary.
Your last paragraph especially reminded me of The Plague by Albert Camus. Beyond its political allegory (and topical subject matter), I read it as an absurdist case for persistent altruism and an approach to everyday life that centers present experiences over hope and despair that map onto the narrative of a broader cause. Both ideas resonated with me later when I read posts in the EA handbook on altruism and scope sensitivity like Nate Soares’ On Caring.
LT🔸
Karma: 39
I haven’t spent a lot of time thinking about this, but I suspect a couple reasons to continue pursuing this contract beyond the present revenue include (1) retaining relationships and a reputation that provides option value for (especially defense-related) future contracts and (2) increasing the likelihood that safer models are used in high-stakes settings, especially ones that could carry some non-negligible AI-related risks. While those are plausible (and plausibly right) lines of reasoning, I’m writing them without taking a stance on specific details that have central importance to their truth (e.g. are Anthropic’s models “safer” than competitors’? Seems quite likely based on reputation, but I’m not well-informed enough to make that claim confidently). If you’re right and the alternative is the use of another lab’s models for the same jobs, and if the article is right that Anthropic’s models are the only ones being used on classified networks, then I don’t think there are good reasons for Anthropic to intentionally cede that space to competitors.
I thought this could be relevant to a few people interested or working in bioethics:
The Bioethics Interest Group is one of several dozen Special Interest Groups that operate out of the Office of Intramural Research at the NIH. Its monthly virtual seminars “provide a discussion forum, consider different views, and present research on complex ethical issues in medical research.” If you are interested in or working in bioethics, I thought you might find it interesting to sign up for its newsletter so that you have the opportunity to read about and consider attending its seminars.
Center on Long-Term Risk, Rethink Priorities, Fish Welfare Initiative, EA Infrastructure Fund
Thank you for the correction! I think what threw me off was the previous video’s impressive 6M views by comparison. I just spent some time looking at examples of how highly successful videos from smaller channels often perform, and I think my perception of the relationship between initial performance and total views was miscalibrated because of how I’ve observed popular videos from well-established channels performing.
I really enjoyed this video! I have one quick note.
Based on the early performance of your first video, I’m a little surprised that this one isn’t on a more similar trajectory in terms of views. While I don’t want to over-speculate,[1] it seems plausible that YouTube’s moderation system may have mischaracterized the title as borderline content, which could have limited its visibility in searches and recommendations.- ^
Turns out I was incorrect about this detail!
- ^
Thank you for writing this! This is a framing I haven’t heard before, and I’m glad to have had the chance to read it.
The thesis of this post appears to be that there are overlooked opportunities in general policy advocacy that could have high leverage in GHD compared to more direct interventions, and that we should pursue them. I like this thesis directionally, but I have a few concerns.
My first is about Nunn’s argument that “since the initiation of antidumping duties requires significant legal capacity, these are typically initiated by wealthier countries and are often against less-developed countries.” Could another explanation be that there are higher rates of dumping in developing countries, potentially due to differences in domestic enforcement capacity or incentives? Some literature suggests that developing countries are generally under less scrutiny for compliance with international trade policy (https://www.chadpbown.com/wp-content/uploads/2019/01/Bown-Hoekman-JWT-2008.pdf), which leads me to doubt that antidumping duties are applied in a way that systematically disadvantages developing countries.
I also have a concern about the claims you make about the importance of consumer power as a lever for improving international labor standards. In the sweatshop example, if higher wages did not come at the expense of less production or fewer jobs, then the change seems clearly good. However, the 400% increase in minimum wages appears to be nominal, since it looks like real minimum wage in Indonesia rose only about 50% from 2000 to 2010 (or 200% from 1990 to 2012) (https://www.researchgate.net/figure/Minimum-Wage-Growth-from-1990-2012-Note-Monthly-nominal-and-real-minimum-wages_fig1_282966999).
In the coffee example, you state that “recent evidence by Dragusanu and Nunn (2018) finds that consumer purchases of fair trade-certified coffee result in increased incomes of FT-certified coffee farmers in Costa Rica.” But the more important question is about incomes across the whole coffee farmer sector (and the laborers those farmers employ), and I don’t know if changes to those groups necessarily followed from changes to the incomes of FT-certified farmers.
Based on the quotes you include, I also share the worry that Nunn’s arguments sometimes risk overemphasizing harms from Western policy. When talking about things like “biased” trade policy, I think it’s pretty important to make clear fact-value distinctions, and I worry that Nunn blurs that line.
I’m sympathetic to the idea that “global aid advocacy is not a substitute for advocacy into good policy by the West,” and I think a more constructive way for it to be framed is as an additional set of (potentially overlooked) items in the arsenal of GHD policy advocates. That said, I have a central concern about whether these generally good policy efforts in developed nations are actually neglected enough to be good candidates for high-value advocacy. As David T noted in another comment, if existing actors already lobby heavily for them, additional EA efforts could have a relatively small marginal impact.
Thank you again for writing this post. While my comment was mostly a criticism, I appreciate what you brought into public discussion on the EA Forum with this post, and I’m glad I had the opportunity to read and comment on it.
This is an amazing speech, and as someone who has to write speeches from time to time, I appreciate the opportunity to hear part of your process laid out in this post.
And another today, this one an exclusive article about the first travel-associated human case in the U.S. connected to the outbreak. My impression is that it isn’t realistically possible for these screwworms to cause a human outbreak in this context, but I think the human cases that will arise from this outbreak could be a compelling part of a broader narrative about its welfare costs and the urgency to control it.
Thanks for writing this! I’m looking forward to reading the rest of the sequence.
I’m curious what approach you’ll take in deciding when to include concrete examples and numbers. On the one hand, examples and statistics from organizations you’ve studied could make your claims more compelling and increase the chance others will apply them in their own decision-making. On the other hand, I recognize there may be privacy concerns with some of the organizations you’ve spoken with, and I know that this isn’t meant to be representative research.
I understand these posts are meant as initial reflections, and I’m glad you’ve chosen to share them. Still, my first reaction was that including more data will likely be important if the outcomes are to be widely taken as practical suggestions for improving governance within EA orgs.
What have you identified as the relevant constraints on the scaling of phage therapies to more clinical settings? I’m not an expert, but I’ve worked on phage biology as a student researcher, and my impression is that the bottleneck on progress for phage therapies isn’t available GMP manufacturing facilities but rather the technical ability to do the necessary tailoring of phage therapies to specific infections at scale and the availability of strong RCT evidence that—with available scalable technologies—we can actually do that.
Since you’re more involved in this field, I realize you may have more current or deep knowledge of the topic, and if you have the capacity, I’d appreciate to hear your perspective on the most relevant bottlenecks on progress in phage therapeutics.
“In Texas cattle country, ranchers brace for flesh-eating screwworms”
Second and third Reuters front-page articles I’ve seen on the issue in the last few months.
I suspect funding and openness to new solutions may increase in response to this growing problem, which could create opportunities for projects like Screwworm-Free Future.
https://www.reuters.com/world/texas-cattle-country-ranchers-brace-flesh-eating-screwworms-2025-08-15/Updated with second article, which I didn’t originally notice.
Thanks for your detailed response! These arguments make sense to me and are valuable to hear, since my background isn’t in animal advocacy or HR. If this does become a full post, it sounds like the results of your research will be valuable to have out there, and I think being explicit about what specific bottlenecks you’ve identified in NGO work will be important part of that.
Is that saturation of great people in nonprofits a result of this nonprofit sector approaching some ceiling/seeing significantly diminishing returns for impact with respect to marginal great people, or is it a result of there being fewer opportunities for marginal impact just at existing orgs? That seems like a potential crux for deciding how to proceed. If it’s the latter, I think it makes sense to ask whether it’s because of a limited number of orgs in general, a limited pool of effective leadership/management for additional orgs, or a limited pool of funding available for additional orgs. I’m assuming you’ve considered and have a good answer for this question; I ended up leaving this comment because I’m curious and because I think it would be valuable to clarify this point if you make a larger post from the content of this take. I also understand that this part of your argument comes from watching the job market, which I think could easily be inscrutable with respect to the contributions of those separate constraints.
That said, I don’t know if that question is a crux of your overall post for me. It seems plausible that the opportunities you refer to lying outside the nonprofit space could be more impactful for the marginal strong candidate regardless, and I’d enjoy reading a longer post that talks about that in more detail.
Several commenters noted that moderate drinking can make casual conversation easier to start and maintain. A related concern I have is that completely abstaining from social drinking might create modest but meaningful barriers to building relationships and networks.
I’m not suggesting that the highest-impact career connections happen at bars, but alcohol does seem to play an important role as an accessory in some personal contexts relevant to professional life. While this doesn’t apply to every setting, I worry that what might appear to be a small restriction on a social behavior could have marginally negative impacts on building both personal and professional relationships, which could aggregate over time and/or across many individuals in communities like EA. Given the value of those professional relationships, I worry that general abstinence from drinking could be moderately costly. I think that’s an important consideration given the popularity of this post.
Two things I’ve read recently influenced this take are Peter Wildeford’s Forum post about “weirdness points” and Nick Cooney’s book Change of Heart, both of which I thoroughly enjoyed and found valuable. I think there’s a broader argument to be made for the value of generally conforming to popular social customs without strong justification to deviate.
That said, I haven’t spent the time to put numbers to this concern, which makes me hesitant about its strength and made me hesitant to post this comment. If I’m wrong, I think a likely reason is that I’m overestimating the social relevance of engaging in moderate social drinking or the extent to which it’s common in relevant contexts. I eventually decided to post this comment because I put credence in the heuristic I proposed in the former paragraph and because I didn’t notice this specific concern raised among the other comments (many of which I found very insightful).
Another thought on this post: Usually when I see a title like the one on this post in the EA Forum, I take it to mean the author is making a recommendation based on principles of EA prioritization (ITN framework/cost-effectiveness/etc.). Andy Masley’s comments indicate that despite his view of the legitimacy of the general arguments behind the recommendation in the post’s title (“you should probably stop drinking”), the arguments don’t meet the bar for EA prioritization. I realize the title is taken from the original blog post, but I worry that its context as a post title on the EA Forum may lead some to assume that the poster is claiming it does meet that bar. I understand this kind of criticism could seem like making a mountain out of a mole hill, but I think it points to important posting norms related to the predictability and reliability of information I would expect many readers consume quickly when reading the Forum homepage.
The best thing I’ve read on it so far is this article by Kelsey Piper.