This is a crosspost of my recent blog post on the specific version of EA I like the most and am excited to defend and promote. I'm working on a much longer post articulating the reasons and justifications for why I chose each point. This post was meant to plant a flag for the basic vision that I'm personally excited about. I'll follow this up with another post going into detail on why I chose the points I did. The first section's written for people less familiar with EA. Most people here can skip it and go to the "Here’s the short version of the EA I’d like to defend:" line! I think getting much more clear and specific definitions and beliefs about EA from more people out there seems pretty important as we continue to get more public attention.


Here’s a very simple model of how society often fails to solve big problems:

  • Basically everyone would like the world to be better in obvious ways, like fewer people dying early or living in poverty. But it’s very easy to get caught up in our own lives and problems.
  • Sticking your neck out to say that something is badly wrong costs you something, and actually trying to fix it costs a lot more.
  • So a lot of ongoing simple and obvious catastrophes go unaddressed, like deadly diseases in poor countries that can be easily prevented with cheap vitamin supplements, or the treatment of animals on factory farms, or the possibility that highly intelligent, capable machines could be uniquely dangerous, because they’re far away, or weird, or hard to fix from scratch, or because nobody gets much credit for working on them.

A lot of our sense of what we should work on, what’s most worthy of good hardworking people, comes from what the people around us admire and notice, which I talk about a lot in my last post on EA. That works fine for lots of big problems, but it also means urgent, terrible problems far away or outside the view of our normal social communities tend to lose out to much smaller ones close to home.

Effective altruism (EA) is basically an attempt to push back against this. It’s a cluster of ideas oriented around using our best evidence and careful reasoning to find the biggest problems that the rest of the world is neglecting and figure out how to actually help in the most quantifiably good way, plus the loose network of people trying to act on them. (If you’ve heard of it lately, it was probably as an insult in a political fight over AI).

Here’s the short version of the EA I’d like to defend:

  • We should take the world’s worst problems seriously, even when they’re far away or strange or unpopular. It is terrible that hundreds of millions of people live in extreme poverty, factory farming is a clear ethical disaster, and the threat that we might lose control of society to AI systems vastly smarter than us needs to be taken more seriously. There are strong moral reasons to devote significant resources (money, time, organizing) to solving these problems, including at some personal cost.
  • We ought to use numbers and explicit cost-benefit analysis where they work… Giving $10,000 to one charity means not giving it to another, so it’s important to compare charities that do similar things. Charities that provide anti-malarial bed nets save more lives per dollar than many other global health charities. Simply counting lives respects people’s basic dignity and freedom, because their personal choices or values don’t come into the equation for whether they deserve help.
  • … but pluralism and respect for other people as free agents are foundational starting points we can’t override. We should among other things reject confident exchange rates between cause areas. We all live surrounded by free people with wildly different moral views, and society works because each of us makes compromises in our own ethics so that everyone can pursue their own vision of the good life. A totalizing utilitarianism ignores this background compromise. A corollary is that choosing which causes to prioritize can’t be reduced to utilitarian math.
  • We should have a strong presumption toward helping others in ways they’re free to turn down (like bed nets, vitamins, and unconditional cash transfers). Utilitarian comparison works best as a way of choosing between interventions that expand people’s freedom and dignity, including their capacity for self-determination. We should also have a strong presumption against biting the bullet of any ethical system that would override people’s decisions about the course of their lives.
  • Ordinary rules about honesty, money, power, and personal/professional boundaries still apply, no matter our perceived ability to rationally analyze our choices. Society’s rules are often good in themselves, and people are too fallible and too swayed by their social scenes and status to safely reason their way past them. No one should feel obligated to justify longstanding social rules and conventions around basic decency from first principles. They often exist for reasons much deeper than we can immediately grasp using reason.
  • EA works best as a cluster of ideas and a loose professional network (a lot like the YIMBY movement) rather than a “community.” When one set of beliefs decides your job, funding, housing and friends, changing your mind or speaking up gets far too costly.

Part of the reason this cluster is so important to me is that it still often works magic and feels like one of the most intellectually vibrant and fruitful scenes in the world. Our work on factory farming is a good example of the good aspects of EA at its best. Taking it seriously is unpopular (many people don’t know how bad it is, or wave away the idea that animals can feel pain) and inconvenient (nobody wants to give up cheap meat), and EA treats it as one of the biggest problems in the world anyway.

Other things EA has managed to do:

  • It has prevented a major war’s worth of deaths, mostly from malaria (GiveWell, a charity evaluator, estimates that the donations it has directed have saved more than 340,000 lives).
  • It helped push the share of American egg-laying hens that are cage-free from under one in ten in 2015 to almost half today.
  • It started building a whole professional field around the potentially massive risks from advanced AI most of a decade before ChatGPT came out. Some EA thinking about AI from years ago still feels shockingly fresh and relevant.

Because of its track record, and how compelling the EA intellectual scene has been, I’ve been publicly associated with and representing effective altruism for the past 5 years, and was a longtime follower of EA debates and ideas in the 10 years before that. It’s shaped my thinking in the same way my other core political and philosophical beliefs have.

Of course, effective altruist ideas have also led to a lot of harm. Sam Bankman-Fried, a major donor who linked himself closely to EA, defrauded depositors of billions of dollars. EA spaces have sometimes tended toward social and intellectual insularity. EA has taken a lot of criticism over the last few years, and while some of it has been in bad faith, some of it has been completely fair.

The same source of EA’s biggest successes is also the source of its biggest risks: if you step outside what the people around you take for granted, that is both how you identify gigantic, obvious problems no one else is working on, but it’s also how people convince themselves that the normal rules don’t apply to them anymore and how they become intellectually insular and isolated. The principles above are the version of EA I’m excited to defend, one that preserves the first and aims to be much better at avoiding the second. This view is rooted in my specific view of ethics and politics, that differs from totalizing utilitarianism, places much more emphasis on human choice and freedom, but still takes many of the trade-offs and choices EA pushes us to consider deeply seriously.

What I’m describing is somewhere between my favorite version of EA as it already exists and a new clear direction I’d like it to move in, learning from its successes and failures so far, and my views on both the value and limits of utilitarian thinking and comparisons. Plenty of people in EA already live by this version, and I’d like it to be easier to point to. This post is meant to plant a basic flag for my high-level principles. I’ve articulated my basic case for EA in much more detail in my last post on it. In future posts I’ll explore in much more detail how I think about EA, what I love about it, the specific version I want to represent, and where I’d like us to reform and change.

56

8
0
7
1
2

Reactions

8
0
7
1
2

More posts like this

Comments6
Sorted by Click to highlight new comments since:

> EA works best as a cluster of ideas and a loose professional network (a lot like the YIMBY movement) rather than a “community.” 

Could you expand on this? The bullet is too short, and I don't know what conclusions to draw here 
 

Hey Andy.   Sorry to bring this up here rather than in one of the many other posts where it could have come up, but there's something that has been bothering me about how animal welfare is discussed in EA.

I often get the impression that animal sentience is treated as an almost indisputable fact. Yet even someone as influential in the broader EA and AI safety world as Eliezer Yudkowsky recently reiterated that his best guess is that chickens aren't sentient. I'm not saying he's right, but surely the fact that someone like him holds this position should give us some reason to be less categorical about the subject?

Given how much emphasis EA places on epistemic integrity and intellectual humility, I find it strange how rarely this uncertainty seems to be acknowledged. Of course, you can believe the evidence strongly favors animal consciousness and advocate for animal welfare accordingly. What bothers me is how often the issue is presented as essentially settled, rather than as a question on which there is still genuine disagreement.

On animal consciousness, the fact that smart people like Eliezer and possibly Jeff Kauffman believe chickens aren't sentient is maybe enough to be ~99% sure chickens are conscious rather than 100%. Speaking for myself (and I expect most EAs would agree), the prior on large animals being conscious due to them having similar brain structures to ours seems really strong. Eliezer seems to disagree due to being extremely confident in a higher order self-modeling theory of consciousness. I find his theory dubious. Even if it were correct, I don't think his theory necessarily implies what he thinks it implies. And even if it really implied chickens are unlikely to be conscious, that wouldn't substantially alter the case for animal welfare.[1]

On the epistemic front, I agree that it's worth being explicit about whatever level of uncertainty we have. But for those who think animals are unlikely to be conscious, I rarely see the same level of humility or prudence on their side -- if there's only a 10% chance chickens are conscious, then EA's cage-free campaign work should be lauded as a valuable long shot, and animal products are still enormously bad in expectation, so they seem worth avoiding.

Edit: On smaller animals like fish, shrimp, and insects, EAs do exhibit more thoughtful discussion about consciousness probability. And Rethink Priorities' Welfare Ranges are conditioned on explicit consciousness probability even for larger animals like chickens.

  1. ^

    Bentham's Bulldog has written a few posts against Eliezer's view of animal consciousness. It's a polemic style, so I won't pretend the author exhibits intellectual humility either, but I agree with the author's arguments.

Thanks Ariel. I agree that uncertainty about animal consciousness doesn't necessarily undermine the case for animal welfare, especially when the potential suffering involved is so large.

But still surprised by the 99% figure, given that beyond Eliezer or Kauffman researchers who study animal consciousness disagree about how conclusive the evidence is. Researchers such as Hakwan Lau and Yoram Gutfreund have questioned whether the available evidence really establishes subjective experience in animals, even mammals and birds. Marian Stamp Dawkins, who has spent decades researching animal welfare, including poultry welfare, published an entire book in 2025 questioning how certain we can be about animal consciousness.

I'm not saying they're right, or even that they represent the majority view. But when people who have spent years working on this question still have serious disagreements about it, I find it hard to see the matter as essentially settled.

And that was really what bothered me about the original post. I understand that the comment about people "waving away" the idea that animals can feel pain may have been aimed at people who simply dismiss the issue without thinking about it. But it can also give the impression that anyone who remains unconvinced must be ignorant or indifferent. I don't think that's fair, especially in a community that values epistemic humility as much as EA does.

getting much more clear and specific definitions and beliefs about EA from more people out there seems pretty important as we continue to get more public attention

Strong agreement. And even before having gotten more public attention, I think this would've been valuable, both for individuals themselves (to get a clearer sense of what exactly they appreciate and decide to follow/make decisions by) and even more for EA overall (to unearth disagreements, allow for constructive debate, and overall advance ideas and principles guiding the movement/community/loose professional network). 

Thanks for having written this up, and I very much look forward to the follow-on posts about each specific point (as well as to similar posts by others describing their respective best versions of EA)!

Curated and popular this week
Relevant opportunities