That 12 people agreed with such a clearly false statement concerns me on what is usually a pretty rational forum. If we include sea fishing and all the insects we kill through spraying, not even close to 99 percent animals we kill are factory farmed. It might even be under half.

Also a line like

“Dwarfs all human problems (including throughout history)” might be true, but it’s extremely uncertain and a great way to turn people off this cause and make people less likely to donate.

over 99 percent of the world doesn’t work on factory farming, and this at least appears to make it sound like you think their work is relatively unimportant.

I think the animal welfare movement still has a big a motivation/​messaging problem which doesn’t help on the donation front. I think there’s a lot to learn from examples like the masterful Lewis Bollard TED talk here which still claims factory farming is the biggest moral problem of our time without unintentionally alienating people.

https://​​youtu.be/​​dvLnIecUNL8?si=VyvYqWjIp701j_MB

I accept that political donations and activism are among the best ways to do good as an individual.

But it is less obvious that EA as an academic discipline and social movement has the analytical frameworks that suit it to politics—we have progress studies and the abundance movement for that.

It is of course necessary for political donations to be analysed as trade offs against donations to other cause areas. And there’s a lot of research that needs doing on the effectiveness of campaign donations and protest movements in achieving expected outcomes. And certain cause areas definitely have issue-specific reasons to do political work.

But I wouldn’t want to see an “EA funds for Democrats” or a “EAs Against Trump” campaign.

I also think this ties back to the broader question I was raising. If a large share of motivated people end up defaulting to earn-to-give not because it’s their best fit, but because pathways into direct impact work are bottlenecked or unclear, that may still point to a structural issue—even if earn-to-give remains net positive in expectation.

So yes, I think the question of “would you be satisfied with a normal job and donating 10%?” is a crucial one. My concern is less about whether that option is impactful in theory, and more about whether the ecosystem is doing enough to help people find durable, high-fit ways to contribute—whether through direct work, earning to give, or something in between.

Changelog: added directorysf (https://​​www.directorysf.com/​​) to the list of places to look for housing. It’s pretty active, but you will need an invite from an existing user to join.

my fave is @Duncan Sabien’s ‘Deadly by Default’

I recently wrote a comment criticizing the Metaculus definition of AGI. I don’t think it is a good starting point for bets as I believe that a system that is clearly not a general intelligence could pass all conditions listed in the question. There is a risk that either the question is resolved based on a non-AGI system that technically fulfills the Metaculus definition (favoring people with shorter timelines), or that the question is never resolved due to its ambiguity (favoring those with longer timelines).

Yeah, totally a contextual call about how to make this point in any given conversation, it can be easy to get bogged down with irrelevant context.

I do think it’s true that utilitarian thought tends to push one towards centralization and central planning, despite the bad track record here. It’s worth engaging with thoughtful critiques of EA vibes on this front.

Salaries are the most basic way our economy does allocation, and one possible “EA government utopia” scenario is one where the government corrects market inefficiencies such that salaries perfectly track “value added to the world.” This is deeply sci-fi of course, but hey why not dream. In such a utopia world, if we really did reach the point where marginal safety researchers are not adding more value than marginal post office workers, salaries would.

My advice for EAs who want to skill up in a neglected area:

1. Get a good mentor who broadly aligns with EA values but already has the skill. I’ve had three of these relationships (with experts in fish welfare, M&E, and agricultural development). In each case, it was symbiotic: I could offer EA knowledge and connections in return for them teaching me their skillset.

2. Book calls with as many relevant people as possible in the field. Shamelessly ask for help, even from people only tangentially related to what you’re trying to learn. The more you snowball contacts, the better your chances of finding a mentor. Also, having lots of people explain the fundamentals to you repeatedly is very useful, as it helps you learn the core ideas of the field.

3. Read a lot. Try to consume a lot of media from the chosen field. You know you’re doing well when you start to notice consistent underlying patterns and premises (again, the core ideas).

4. Start applying it before you’re ready. Look for small projects you can use as practice for the new skill.

In general, when learning a new skill, Andrea Gunn’s talk on training leaders offers a lot of good insight. I also made a one-page summary of her talk.

I have historically been able to do this upskilling as a side project to my existing job.

Also failures trying to do really outlandish things like bribing Congresspeople to endorse Jim Mattis as a centrist candidate in the 2024 US Presidential Election are likely to backfire in more spectacular ways than (say) providing malaria nets for a region with falling malaria or losing a court case against a factory farming conglomerate. That said, this criticism does apply to some other things EAs are interested in, particularly actions purportedly addressing x-risks.

Executive summary: This post explains what the MATS Summer 2026 program is like in practice and argues that it is a good fit for self-directed researchers motivated to work on reducing risks from unaligned AI, while being a poor fit for those seeking structured coursework or credentials.

Key points:

1. The author states that MATS aims to identify and train talented people to work on reducing risks from unaligned AI by providing mentorship, funding, and community rather than a formal curriculum.

2. Since 2021, MATS has supported over 500 researchers with 100+ mentors, resulting in 160+ papers, 7,800+ citations, and roughly 80% of alumni working directly in AI safety or security.

3. Fellows receive a 12-week fully funded research phase with mentorship, a $15,000 stipend, a $12,000 compute budget, housing, and access to seminars and the AI safety community.

4. Most fellows work on independent research projects with flexible mentorship styles and support from dedicated research managers, culminating in milestones like a research plan and poster presentation.

5. Around 75% of fellows continue into a 6–12 month extension phase, which accounts for roughly 60% of total scholar-time and expects high research autonomy.

6. The program is best suited to applicants comfortable with ambiguity, independent research, and revising ideas, and is not intended for those seeking structured assignments, credentials, or tightly defined tasks.

This comment was auto-generated by the EA Forum Team. Feel free to point out issues with this summary by replying to the comment, and contact us if you have feedback.

Executive summary: The post offers a reflective overview of artificial wisdom as a nascent, fragmented research area, arguing that despite severe challenges in taxonomy, funding, careers, and conceptual coherence, deliberate field-building and strategic positioning could allow it to mature into a critical component of AI alignment and long-term human flourishing.

Key points:

1. The author argues that artificial wisdom research is highly fragmented due to inconsistent terminology, poor discoverability, and the absence of dedicated conferences, journals, or institutional homes.

2. Compared to established AI safety subfields like interpretability, artificial wisdom faces a comparative disadvantage in funding, recognition, and career infrastructure because its questions are long-term, philosophical, and harder to operationalize.

3. The post claims that community-building efforts such as fellowships, research incubators, and collaborative platforms are necessary to prevent the field from remaining marginal.

4. The author contrasts independent research with doctoral programs, noting tradeoffs between intellectual freedom, financial security, legitimacy, and institutional constraint.

5. Researchers are advised to strategically frame artificial wisdom work using established concepts like coherent extrapolated volition or the long reflection to improve funding accessibility and legitimacy.

6. The post describes major conceptual divides in the field, including outcome-oriented versus process-oriented definitions of wisdom and top-down versus bottom-up architectural approaches, while noting broad agreement on the importance of meta-ethical reasoning.

This comment was auto-generated by the EA Forum Team. Feel free to point out issues with this summary by replying to the comment, and contact us if you have feedback.

Yes, and also:

(1) PR benefits: Being fit makes people doing public outreach to raise awareness of X-risks more credible, persuasive, charismatic, & energetic, and better able to handle the physical & mental stresses of public engagement.

(2) Survival benefits given global catastrophic (if not X-risk) scenarios: As every serious prepper & survivalist knows, physical fitness is a crucial element of surviving in case of ‘SHTF’ or ‘TEOTWAWKI’ scenarios. If infrastructure fails (e.g. internet fails, electricity/​water/​gas fails, or supply chains fail), it’s important to be able to run, hide, fight, carry, etc.

And, if we don’t look physically prepared for these survivalist scenarios, it may be harder for normal people to take us seriously when we talk about X-risks. (On the principle that if we took our own worst-case doom scenarios seriously, we’d also be getting prepped for less-than-worst-case scenarios—which are logically both more likely, and more imminent).

This summary is accurate.

Hi Jenny, very interesting, thank you. What was the response of CG to your report, and do you know if they are planning to invest more resources towards this potential cause area?

I like that this question allows to some extent to incorporate discussions about EA’s blindspots in ‘messy’ processes like politics /​ institutional change etc. by asking what to do in such areas.

1. Depends on what you count as meaningful earning potential.
   
   One of the big ideas that I take from the old days of effective altruism, is that strategically donating 10% of the median US salary can save more lives than becoming a doctor in the US over one’s career.
   
   Same logic applies to animal welfare, catastrophic risk reduction, and other priorities.
   

2. A different question is would you be satisfied with having a normal job and donating 10% (or whatever % makes sense in your situation)?
   

I was wondering if anyone was going to mention that. There was a lot of media buzz about whether the events of the show could really happen at the time of its airing. This piece by Yale is supposed to sound reassuring, but it just… doesn’t. :/​

Among other things, the natural-atrocity take on zombies is what got me in love with the TV series; depressed by them but interested in this new aesthetic of dangerous nature globally killing human civilization, think overgrown moss on broken subways. I can indeed see things like it motivating EA people to prevent such things, very much involved with visions of such a world. 🪸

Max Tegmark explains it best I think. Very clear and compelling and you don’t need any technical background to understand what he’s saying.

I believe his third or maybe it was second appearance on Lex Fridman’s podcast where I first heard his strongest arguments, although those are quite long with extraneous content, here is a version that is just the arguments. His solutions are somewhat specific, but overall his explanation is very good I think:

To my knowledge, there are a few (not actually that many) existing antifungals, but as I commented above, they mostly aren’t very good, and in several deadly fungal infections they are almost pointless.

Also, when a new fungal pathogen comes out, it might be harmless, or it might be big trouble, nobody can predict that. A good example I’ve seen mentioned a few times is Candida Auris (pretty serious and often deadly fungal infection) that emerged in 2009 independently in several regions of the world, pretty much out of nowhere. And the scary thing is that it was drug-resistant from the start! I think researchers aren’t quite sure why it emerged, but it could be related to climate change.