Seems worth establishing the fact that bad actors exist, will try to join our community, and engage in this pattern of almost plausibly deniable shamelessly bad behavior. I think EAs often have a mental block around admitting that in most of the world, lying is a cheap and effective strategy for personal gain; I think we make wrong judgments because we’re missing this key fact about how the world works. I think we should generalize from this incident, and having a clear record is helpful for doing so.
BenHoffman
Matching-donation fundraisers can be harmfully dishonest
The wording is just to make it a stronger, better commitment.
How does it do that? Is that effect stable under conditions where people don’t see the pledge as a binding promise?
Look out in the world and you’ll see lots of people excited about things that don’t work or don’t do what they say they’ll do. Anyone can say they’re evidence-backed etc. On outside view, if you only spend a few minutes on your donations each year, how much of the optimization pressure influencing your donations should you expect was marketing skill on the part of the recipient or their patron, vs selecting for actual impact?
Do you disagree with the first bullet point? Or do you disagree with the second? Or do you disagree that they jointly imply something like the bit you quoted?
That’s good to hear. But I didn’t think you were saying that criticism is generally harmful—I thought you were saying that failing to check in with GWWC first is harmful in expectation. If so, I’m curious what the most important scenarios are in which it could cause harm to start this sort of conversation in public rather than in private. If not, when do you think this advice does help?
It additionally seemed like you thought that this advice should be applied, not just to criticism of GWWC’s own conduct, but to criticism of the idea of the pledge itself—which is already public, and not entirely specific to GWWC, as organizations like The Life You Can Save and REG promote similar pledges. I got this impression because Alyssa’s post is limited to discussion of the public pledge itself.
I don’t think that Ben Todd is proposing (5). I think he’s proposing (4), and that this proposed norm would effectively be a tax on criticism. Taxes aren’t as costly as bans, and can be good if they pay for something good enough, but in this case I don’t think it’s worth it.
In particular, applying journalistic standards to criticism of, but not praise of, EA orgs’ behavior seems like a weird position to take if what you’re interested in is improving the quality of public information.
For what it’s worth, your comment helped me clarify my position, and I wish I’d been able to express myself that clearly earlier.
Also, somewhat embarrassingly, I am also Benquo (I think I accidentally signed up once via mobile, forgot, and signed up again via desktop.) Hopefully I’ll remember to just use this login going forward.
My thoughts on this are too long for a comment, but I’ve written them up here—posting a link in the spirit of making this forum post a comprehensive roundup: http://benjaminrosshoffman.com/honesty-and-perjury/
You’ve described a project at a fairly high level of abstraction. You’ve already put 20-40 hours in, so your research has already likely taken some specific directions. Sharing a brief summary of this would help people with compatible approaches who think you’re doing potentially overlapping work notice that they should reach out to you. It would also help save the time of those who aren’t members of that group.
Peter just suggested you mention more details about the project, in the comments. Daniel did too. As a reader, I would have benefited if you’d replied by giving them details about the project. I expect there are more readers like me, who might reach out if a project seemed like it was going in an interesting direction (even if not my preferred direction), but not without such a specific reason to think it’s worth their time.
If there are specific reasons for discretion, of course, you can say so.
I haven’t yet seen a formal approach I find satisfying and compelling for questions like “How should I behave when I perceive a significant risk that I’m badly misguided in a fundamental way?”
Seems like the obvious thing would be to frontload testing your hypotheses, try things that break quickly and perceptibly if a key belief is wrong, minimize the extent to which you try to control the behavior of other agents in ways other than sharing information, and share resources when you happen to be extraordinarily lucky. In other words, behave like you’d like other agents who might be badly misguided in a fundamental way to behave.
Yes
Depending on the circumstances, a focus on preserving EA as a movement and avoiding disruptions to existing top philanthropic opportunities may miss the most important opportunities. My guess is that we’ll do better asking questions like:
What types of disruptions might hamper our ability to coordinate with one another and outsiders to improve the world or mitigate emerging problems? (Different sub-problems may demand very different solutions.)
How can we solve these problems in a way that works for EA and other individuals and groups trying to do good? (We should try to generate solutions that transfer well, not just solve the problem for ourselves.)
Who else is already working on similar problems RE making global cooperation more robust to war or other likely disruptive events? What can we do to help them or benefit from their help?
What disruptions are EAs especially well placed to mitigate?
Which interventions are likely to be most important in the event of various disruptions?
This seems like it describes the most relevant considerations when thinking about how the pledge directly affects your own future actions. I think there’s another angle worth considering, and that’s the pledge as a report about your likely future actions.
You might be reluctant to take the pledge, not just out of a worry that you’ll bind your future self to a wrong decision, but out of a worry that if your future self acts differently, people in the meantime will have made decisions based on your assurance. It’s quite difficult to model others well enough to figure out whether they’ll take the optimal action as you see it, but potentially easy to decide whether to believe their promises. This makes coordination easier.
A couple of years ago, a friend was considering a relocation that improved their expected lifetime impact substantially. Moving would potentially have put their personal finances under strain, so I offered to lend them a few thousand dollars if money should happen to be tight for a while. They found this offer sufficiently reassuring that they were happy to go ahead with the move without delay. I felt that the offer was morally binding on me barring severe unforeseen circumstances, but the point of my promise was neither to coercively bind my future self, nor simply to determine my future self’s course of action by establishing some momentum. The point was to accurately report my future willingness to lend to my friend, with high confidence.
If it had turned out to be somewhat harder than anticipated to lend my friend the money, I would have considered myself obliged to work hard to figure out a solution. I don’t think this was especially related to the fact that the behavior I was making a promise about was mine. If I ever make an assurance to someone, and they end up harming themselves because it turned out to be a false assurance, I consider myself at least somewhat obliged to try to make them whole.
Giving What We Can, for example, uses the number of people who have taken the pledge as a measurement of their impact. Giving What We Can itself and potential GWWC donors make decisions about whether promoting the pledge is a good use of resources, based on both the observed behavior of pledgers, and some beliefs about how serious pledgers’ intent is. When considering publicly pledging, you should consider not just its effect on you, but that you’re either providing accurate information or misinformation to those who are paying attention.
For this reason, I think that serious pledges should not be entered into unless the pledge is either easy (i.e. you predict with high confidence that you would do the pledged behavior anyway) or very important (you predict that taking the pledge gives you options much more valuable than the ones that might otherwise be available). An example of an easy pledge would be assuring a future houseguest that coffee will be available (if you regularly stock coffee). An example of a very important pledge might be marriage, in which by promising to stick with someone, you get them to promise the same to you—though many people delay getting married until the promise feels easy as well.
In principle, if there’s unmet demand for these things, then it’s a great idea to set up such funds. Overall this infrastructure seems plausibly helpful.
But I’m confused about why, if this is a good idea, Open Phil hasn’t already funded it. I wouldn’t make such a claim about any possible fund set up in this way—that way leads to playing the Defectbot strategy in the iterated prisoner’s dilemma. But in this particular case, I’d expect Open Phil to have much more reason than outside donors do to trust Elie’s, Lewis’s, and Nick’s judgment and value-alignment. Though per Kerry’s “minimum viable product” comment below, perhaps this info asymmetry argument will be less true in the future.
I suspect that Open Phil is actually making a mistake by not empowering individuals more to make unaccountable discretionary decisions, so this is seems good to try in its current form anyhow. I weakly expect it to outperform just giving the money to Open Phil or the GiveWell top charities. I’m looking forward to seeing what happens.
I want to encourage this person to:
Write about what you’ve learned doing direct work that might be relevant to EAs.
Reach out to me if I can be helpful with this in any way.
Keep doing the good work you know how to do, if you don’t see any better options.
Stay alert for high-leverage opportunities to do more, including opportunities you can see and other EAs can’t, where additional funding or people or expertise that EAs might have would be helpful.
so much!
I wish to register my emphatic partial agreement with much of this one, though I do still identify as EA, and have also talked with many people who are quite curious and interested in getting value from learning about new perspectives.
Relevant resources:
The Open Philanthropy Project’s Shallow Investigations provide nice template examples.
The Neglected Virtue of Scholarship
Scholarship: How to Do It Efficiently
I’m fairly new to the EA Forum, maybe someone who’s been here longer knows of other resources on this site.
You can’t arbitrarily get more expected returns for taking on more risk, except possibly with complicated financial instruments and trading.
Paul Christiano suggests leveraged ETFs. There’s also buying stocks on margin, which is not terribly hard to set up.
There’s an important difference between repeatedly making errors, jumping to conclusions, or being attached to a preconceived notion (all of which which I’ve personally done in front of Carl plenty of times), and the sort of behavior described in the OP, which seems more like intentional misrepresentation for the sake of climbing a social status gradient.