I agree that domain knowledge was a real problem here.

I don't see, however, how this evidences the idea that earlier grantmaker involvement would mitigate this. The grantmaker was involved on both sides of the transaction, presumably from an early stage, so if anything it should update you in the other direction.

I don't think conflict of interest norms are mainly a product of states needing legibility, in the Seeing Like a State sense (although I haven't read that book). The principle that no one should be a judge in their own cause is centuries old. High-trust communities built on expertise, like academic peer review, adopted conflict rules for themselves rather than having them imposed.

I take your points, which I think are broadly true - people and organisations use weak proxy metrics a lot, sometimes for the wrong reasons, and we shouldn't adopt norms for their own sake.

At the same time, I think conflicts of interest are a norm we should take seriously, and for substantive reasons. Conflicts of interest genuinely get in the way of truth-seeking, by making people less objective and making it harder for them to decide what to do and say. They therefore disadvantage the ecosystem, organisations and individuals, including those individuals in conflicted positions.

Some examples:

• Funder A comes to person X for a reference or assessment of a charity. Person X sits on the Board of that charity and knows something very bad that happened internally, but is bound by confidentiality. This puts person X in a tough situation.
• Funder B has a Board seat at charity Z. Charity Z decides to apply to Funder B for a large grant and is very enthusiastic about it. Funder B actually thinks it's a bad grant. They are forced to decide between the best funding decision they can make, and disappointing the Board and staff of charity Z.

I accept your point (and made the same one myself) that conflicts are almost inevitable in small ecosystems, but my response to that is that we mitigate this by acknowledging, declaring and managing the conflict. So, for example, person X can give a reference after saying "as a trustee, I am clearly conflicted and not able to give my fullest reference, but you can have it anyway." Funder B can say "as I am conflicted, I will ask an independent colleague to assess this grant."

This also answers your point that the purpose of Claire being there may have been to help ensure the money was spent well. That's an argument for her sitting on the Board, but not for her being the person at OP who assessed the grant. She could have overseen the spending as a trustee while a colleague assessed the grant, without any noticeable loss to anyone involved.

Even if you weight these norms less than I do, the cost of mitigation in this case was close to zero - namely one colleague assessing the grant instead of her. I can't see that much was gained by skipping it, and I think this applies across almost all conflicts of interest - the cost of mitigation is small and the dangers of no mitigation much larger.

Thanks for this, and in particular for writing a strongly argued disagreement, which takes courage.

I didn't particularly want to relitigate the Wytham Abbey debate but I guess I did mention it, so fair enough.

This may be me retrospectively changing my motives, but my objections to Wytham Abbey were (I think) threefold:

1. The building seemed incredibly poorly suited to the proposed use case. Wrong layout/rooms, subject to heavy restrictions as a historic building, substandard catering facilities. I would have been less bothered by them buying a fit-for-purpose conference centre (I think). I think this point of view is one reason why it was used so sporadically, and I vaguely remember people who went there said it was a crap place to hold retreats etc.

2. The conflict of interest in the grant decision.

3. The optics.

I think you argue quite coherently against the optics and this is the one I am least confident in and feel least bothered about today. I agree that EA has commendably defied the optics to do important things.

Your own footnote says it "may have been a bad investment if it were underutilised or if it didn't deliver significant value over hiring external venues". I think it was. It was used sporadically, was hugely controversial, became a meme of EA's excesses and failures, and EV sold it in 2025 for £6m, having bought it for £14.9m in 2022.

On the conflict of interest, I linked to that comment because I followed up with specific (and, in my view, reasonable) questions that went unaddressed. I raised this issue, Claire linked to a comment which did not address my concerns, I clarified my concerns and my follow-up questions weren't answered.

I still emphatically disagree with you on this. There is, in my view, no acceptable reason why that grant was assessed by someone who was so obviously conflicted, when it would have been so easy to mitigate this conflict by having almost literally anyone else at OP assess the grant. (Bear in mind that the widely quoted and accepted standard is an actual or perceived conflict of interest.)

I can also say that I have never heard of, nor seen any evidence of, large grantmakers* taking a Board seat and then having that same person assess the specific grants to that organisation, although that doesn't necessarily mean that this doesn't happen. For the record, even if the practice is widespread, I would still disagree with it on principle, but I would update in your direction.

You and I can obviously have differences of opinion, but you should consider carefully that there is widespread agreement in society, charitable regulation etc. that conflicts of interest are important and should be both acknowledged and mitigated, so the onus is on you to justify your opinion that these processes miss "real virtue or integrity", as there are very strong and widespread norms in the other direction.

I think you're right that managing this conflict wasn't legally required, as far as I can tell. My point is one that I think you gesture at in your footnote 3, about breaches of trust with the public: the norm exists for exactly this situation, because high trust inside a group doesn't substitute for the confidence of people outside it.

(I don't have any opinion on Claire's actual virtue or integrity and, if you do, this may lie behind your reaction. My opinion is that this specific grant was terrible governance, but that can be a mistake as easily as it can be malpractice.)

On your specific criticisms:
a) No, she did not give a reasonable explanation for why she assessed a high value and highly controversial grant to an organisation on whose Board she sat, when easy mitigation was at hand. She also did not provide any account that I have seen regarding whether this conflict was acknowledged, disclosed or managed, as ordinary governance would seem to demand.
b) No, it didn't, because the principles would have been served very well (probably equally as well) by someone else competent assessing the grant, without nearly so many concerns about conflicts of interest.
c) I guess I have answered this with points a) and b).
d) I can see that this might seem arrogant. My intention was to show that serious conflicts of interest existed, were apparently not mitigated and then were waved away without adequate explanation, and I think this comment thread demonstrates that.

Finally, I do agree that FTX imploding gave lots of ammunition to people who were wrong in expectation to say "I told you so". FTX being a fraud doesn't mean that everyone who objected to FTX-related things was right at the time, including me. Equally, the fact that Wytham Abbey failed doesn't prove it was a bad grant in expectation. But my first objection was about the building itself, which was knowable before the purchase, and that is how it played out in reality.

----------------------

*I am not sure how to footnote in these comments, but just to note, I said "large grantmakers" because this can be very difficult in small grantmakers, if you want to take Board seats. For example, I hold a Board seat at Lafiya as part of my grantmaking relationship to them. I joined the Board in December 2025, after our Fund's grant to them, and might have decided on a further grant to them after taking the seat (as it happens, I haven't done this, but I could have).

Taking that Board seat was an experiment, which I do not intend to repeat. I am also now asking for two second opinions on any future grants to Lafiya, to mitigate the conflict of interest. I just wouldn't want this to come out and for me to seem hypocritical.

Also, for the record, if Ultra Philanthropy (currently 1.8 FTE on payroll plus some contractors) can take steps to mitigate conflicts of interest over six-figure grants, it makes me even more insistent that a multibillion-dollar foundation with hundreds of staff can do the same, when making controversial eight-figure grant decisions.

Thanks Alex - this is an interesting take.

In general, gift acceptance policies are useful, if only to stop the charity accepting unwanted gifts (e.g. a crumbling building that costs a lot to maintain).

I am not sure if they solve the specific issue you have in mind here, though. An unrestricted gift of $100m would pass most policies I have seen. Or were you thinking to include some sort of "undue influence" clause?

I wonder if this protection relies more on a) leaders have the strength and integrity to reject a major donor's advice; but also b) a diverse funding landscape, where even a $100m donor has less clout because they are not a disproportionate part of the funding ecosystem (see here: https://open.substack.com/pub/fundinganthropalypse/p/two-coefficient-givings-beat-one?utm_source=share&utm_medium=android&r=4b7xoz)

I love this Patrick, and it sounds like I am belatedly trying to create the set up you found with the Integral Altruists.

Thanks Huw, I agree with lots of this, in particular that a lot of FTX mistakes were decent people making bad decisions, rather than malicious actors. I think the latter is more likely than the former, but both are possible.

I found Chana's ideas on epistemically rigorous gossip helpful to think about checks and balances, provided that someone has some semi-formal role to look into them and take action. This is why I asked Sjir to help - if there are rumours circling about Ultra, there needs to be someone who is empowered to look into them and report back. (Chana's post: https://forum.effectivealtruism.org/posts/KE4Ga3zHQsczooQi7/correctly-calibrated-trust#:~:text=epistemically%20rigorous%20gossip)

It's interesting that you mention Tara because one of my reflections after FTX was "how is it possible that so many people knew this guy was an outright liar and yet no one stopped him?"

I think this came down to two things:

1. Lots of people chose to keep their concerns to themselves, because they were afraid of retribution, or thought someone else was dealing with it. I think robust whistleblowing protections mitigate this somewhat, by giving them protection and placing a report formally on the record. It would also help if people made an internal promise to themselves to act on concerns if they have them, rather than just moving on for the sake of a quiet life (although I recognise that this is hard).

2. As I understand it, some people with power did have a fuller picture of Sam's behaviour and decided not to do anything about it. This is partially solved by the whistleblowing element (it puts a complaint on the record and makes it harder to ignore), and partly by having people whose role it is to investigate and adjudicate these things.

Obviously one could still simply ignore or fire one's ethics advisors, but I feel like Sjir would resign and then write a forum post if I did that, and that would affect my ability to keep acting badly.

I do agree though that SBF was hard to stop, because oathbreakers are very hard to contain.

In sum, I hope that some combination of personal accountability/conscience and institutional checks and balances can help us navigate this one much better.

This is a really interesting and helpful addition to the discussion.

I think I agree with you that it will probably help, if doing impactful things also accrues money/status/power, in that it will you make highly motivated to do more impactful things and make you feel good for doing them.

I think, though, that there are key moments where the impactful thing diverges from the money/status/power thing, and in those cases, I would want people to be clear that their primary goal would be impact. So I don't endorse the idea that we should say "well we are biologically programmed to pursue money and status, so let's not worry about it", if that's what you are saying.

I think this is a valid concern and it obviously the diamond is in some ways a form a virtue signalling. That effect is likely to increase as EA becomes cool again, which it seems to be. (It's worth saying that when GWWC launched this, EA was very uncool, and adding the diamond felt like quite a big risk, at least to me. That isn't the case now, though.)

I think the main idea behind the diamond is that telling people that you donate can inspire others to do the same, or to talk to you about giving. I am not sure how effective this is, or how to trade it off against the concerns you raise.

I guess I also think that EA is quite money and donation-centric, albeit that this isn't the sum total of what we do, but it is something we have done very well historically.

Thanks Grace.

These indicators seem like good ones to me. I particularly dislike the "well they [less powerful people] just don't understand what we are doing" thing.

To close the loop on this: you were right, it was an oversight on our part, and we've now taken mental health off GiveWell's profile, with their agreement.

I wasn't offended, but thanks for following up :)