The impact of the FTX scandal on EA is starting to hit the news. The coverage in this NY Times article seems fair to me. I also think that the FTX Future Fund leadership decision to jointly resign both was the right thing to do, and comes across that way in the article. Will MacAskill, I think, is continuing to show leadership interfacing with the media - it's a big transition from his book tour not long ago to giving quotes to the press about FTX's implosion.
The article focuses on the impact this has had on EA:
[The collapse of FTX] has also dealt a significant blow to the corner of philanthropy known as effective altruism, a philosophy that advocates applying data and evidence to doing the most good for the many and that is deeply tied to Mr. Bankman-Fried, one of its leading proponents and donors. Now nonprofits are scrambling to replace millions in grant commitments from Mr. Bankman-Fried’s charitable vehicles, and members of the effective altruism community are asking themselves whether they might have helped burnish his reputation...
... For a relatively young movement that was already wrestling over its growth and focus, such a high-profile scandal implicating one of the group’s most famous proponents represents a significant setback.
The article mentions the FTX Future Fund joint resignation, focusing on the grants that will not be able to be honored and what those might have helped.
The article talks about Will MacAskill inspiring SBF to switch his career plans to pursue earning to give, but doesn't try to blame the fraud on utilitarianism or on EA. This is my opinion, but I'm just confused by people's eagerness to blame this on utilitarianism or the EA movement. The common-sense American lens to view these sorts of outcomes is a framework of personal responsibility. If SBF committed fraud, that is indicative of a problem with his personal character, not the moral philosophy he claims to subscribe to.
His connection to the movement in fact predates the vast fortune he won and lost in the cryptocurrency field. Over lunch a decade ago while he was still in college, Mr. Bankman-Fried told Mr. MacAskill, the philosopher, that he wanted to work on animal-welfare issues. Mr. MacAskill suggested the young man could do more good earning large sums of money and donating the bulk of it to good causes instead.
Mr. Bankman-Fried went into finance with the stated intention of making a fortune that he could then give away. In an interview with The New York Times last month about effective altruism, Mr. Bankman-Fried said he planned to give away a vast majority of his fortune in the next 10 to 20 years to effective altruist causes. He did not respond to a request for comment for this article.
Contrary to my expectation, the article was pretty straightforward in describing the global health/longtermism aspects of EA:
Effective altruism focuses on the question of how individuals can do as much good as possible with the money and time available to them. Historically, the community focused on low-cost medical interventions, such as insecticide-treated bed nets to prevent mosquitoes from giving people malaria.
More recently many members of the movement have focused on issues that could have a greater impact on the future, like pandemic prevention and nuclear nonproliferation as well as preventing artificial intelligence from running amok and sending people to distant planets to increase our chances of survival as a species.
Probably the most critical aspect of the article was this:
Benjamin Soskis, senior research associate in the Center on Nonprofits and Philanthropy at the Urban Institute, said that the issues raised by Mr. Bankman-Fried’s reversal of fortune acted as a “distorted fun-house mirror of a lot of the problems with contemporary philanthropy,” in which very young donors control increasingly enormous fortunes.
“They gain legitimation from their status as philanthropists, and there’s a huge amount of incentive to allow them to call the shots and gain prominence as long as the money is flowing,” Mr. Soskis said.
But even this focuses on the same problems of the purchasing of status with philanthropy that we ourselves are wrestling with right now.
Edit: See Aaron Chang's comment for what he sees as the most glaring issue pointed out by the article - "loose norms around board of directors and conflicts of interests between funding orgs and grantees"
I expect that other articles may take a harder look at EA. But I was heartened to see that in this case at least, the author, Nicholas Kulish, seems to be treating EA like what I understand it to be -- a lot of people earnestly trying to figure out how to make the world a better place, and trying to find the most ethical way to navigate a disaster.
I think EAs could stand to learn something from non-EAs here, about how not to blame the victim even when the victim is you.
I agree the victim-perpetrator is an important lens through which to view this saga. But, I also think that an investor-investee framing is another important one; a framing that has different prescriptions for what lessons to take away, and what to do next. The EA community staked easily a billion dollars worth of its assets (in focus, time, reputation, etc.), and ended up losing it all. I think it's crucial to reflect on whether the extent of our due diligence and risk management was commensurate with the size of EA's bet.
As someone who has spent years spreading the message that humans are very prone to self-serving biases (hopefully this is an acceptable paraphrase of some complex ideas!), I've personally been surprised to see your many posts in the forum right now which seem to confidently assert that the outcome was both unforeseeable and unrelated to rationalist ideas (therefore making EAs including yourself purely victims, rather than potentially also causal agents here).
To me, there seems a really plausible path from ideas about the extreme urgency of AI alignment research & the importance of taking "extreme" personal agency (relative to existing social norms) to a group of people taking on extreme risks with a lot of urgency and high personal agency to raise funds for AI alignment research.
I have no connection to any of the people involved and no way to know whether it's what happened in this case, I'm just saying that it seems like a plausible path to what happened here given the publicly available information, and I'm curious whether that's something you've considered.
The last paragraphs in the article itself point to the most glaring issue IMO-loose norms around board of directors and conflicts of interests (COIs) between funding orgs and grantees. The author presents it in a way that it's self evident the boards were not constructed in a way to be sufficiently independent / objective, and having substantial overlap between the foundation board and the boards of the largest grantees can lead to hazards. These are common industry issues in corporate oversight, curious what policies there are among EA orgs to decrease COIs.
"A significant share of the grants went to groups focused on building the effective altruist movement rather than organizations working directly on its causes. Many of those groups had ties to Mr. Bankman-Fried’s own team of advisers. The largest single grant listed on the Future Fund website was $15 million to a group called Longview, which according to its website counts the philosopher Mr. MacAskill and the chief executive of the FTX Foundation, Nick Beckstead, among its own advisers.
The second-largest grant, in the amount of $13.9 million, went to the Center for Effective Altruism. Mr. MacAskill was a founder of the center. Both Mr. Beckstead and Mr. MacAskill are on the group’s board of trustees, with Mr. MacAskill serving as the chair of the United Kingdom board and Mr. Beckstead as the chair of the U.S. subsidiary.
The FTX Foundation itself had little to no oversight beyond Mr. Bankman-Fried’s close coterie of collaborators. According to its website, the board of the FTX Foundation comprised Caroline Ellison, the head of Alameda Research, the hedge fund Mr. Bankman-Fried founded; Gary Wang, the chief technology officer of FTX; and Nishad Singh, director of engineering at FTX."
I think the point of most non-profit boards is to ensure that donor funds are used effectively to advance the organization's charitable mission. If that's the case, then having donor representation on the board seems appropriate. Why would this represent a conflict of interest? My impression is that this is quite common amongst non-profits and is not considered problematic. (Note that Holden is on ARC's board.)
I'm also not sure this what the NYT author is objecting to. I think they would be equally unhappy with SBF claiming to have donated a lot, but it secretly went to a DAF he controlled that he could potentially use to have influence later. The problem is more like trying to claim credit for good works despite not having actually given up the influence yet, not a COI issue.
(I don't think it's plausible to call "I gave my money to foundation or DAF, and then I make 100% of the calls about how the foundation donates" a COI issue. )
I don't see how this follows.
It is indeed very normal to have one or more donors on the board of a nonprofit. But FTX the for-profit organization did in fact have different interests than the FTX Foundation. For example, it was in the FTX Foundation's interest to not make promises to grantees that it could not honor. It was also in the Foundation's broader interest to safeguard the reputation of the broader EA community. Both of these interests were in fact jeopardized by FTX's actions. It's reasonable to argue whether an average nonprofit board, exercising average levels of diligence, would have pushed back more on FTX. Maybe not. But that there was a conflict is clear ex post, and I think it's very plausible that having the FTXF Board have one or more fiduciaries that were not simultaneously beholden to FTX would have caused the org to be a lot more cautious in a variety of ways. At least, that's the normal corporate governance theory.
OpenPhil has a majority of board members (3/5) who aren't the source of funds (Moskovitz and Tuna, who are the other 2). As I understand it, they also have a few $B under their direct independent legal control. The fact that FTX Foundation didn't secure any assets independently this way is a massive failure (for the world, EA, and FTX creditors).
Would be good to see proof of this.
Were there significant assets in an independently-controlled FTX Foundation we would be in a much better position now even from the point of view of wanting (or being compelled) to use the money to pay back FTX creditors.
My original comment mentions 'can lead to hazards' due to not knowing the details of what went down/being relatively new (couple months) to the EA space. However subsequent comments of individual regrantors needing to self 'reinvent the wheel' around COI policies (btw, I think Linch came to a good equilibrium for his particular situation) and other concerns raised in the past I think confirm the hunch that this is an area of improvement.
Yes, that is one function of boards. Funders also ideally function objectively when choosing who to fund. Not so hypothetical example: Foundation can fund grantee A or B. Foundation employee making the decision is board member/trustee of B with fiduciary duties to B. Should foundation employee fund A or B? Conflict. Employee should probably recuse themselves from the decision making process and document. If employee was a board observer on B with no decision making power at B, it might be ok, but might still a good idea to recuse themselves from the fund distribution process if want to be safe/ doubly above board.
Most folks who have been required to go through COI learning modules for work can attest to how fun they are <sarcasm>, but I guess I'm seeing the importance now. There are general good practices because of problems that arise with 'self serving' and I think the policies of well known foundations speak for themselves and are instructive if anyone wants to dive deeper. I included COI policies from MacArthur Foundation and Bill and Melinda Gates Foundation and they are similar to policies in the government contracting and healthcare spaces which are more regulated. For reference, MacArthur states that "a grant is material to an entity when the amount of the grant is in excess of five percent (5%) of the revenue of the entity." I suspect that is the case in several EA orgs, so may be good to at least revisit their COI policies in light of recent events.
One key thing is a person's degree of decision making power in the funder org vs grantee org and situations of recusal when a COI is identified. I think what the NYT author is getting at by just stating the facts and offering little commentary is that how things have operated does not pass the sniff test for having functioning COI policies (from outside looking in). Happy to update thinking if there is evidence to the contrary that these COIs were brought up and CEA/FTX Foundation worked through them (general best practice, see 'record and reporting' section of the MacArthur Foundation policy below). I.e. credit to EA Infrastructure Fund, from a quick search I noticed they mentioned COIs under "Why you might choose not to donate to this fund" so at least some things are working: " You have concerns about conflicts of interest or grantmaker independence: All members of the fund management team are actively involved in the effective altruism community and have both professional and personal relationships with many of those working at meta organizations that may receive grants from this fund. Nick Beckstead (an advisor to the Fund) is a member of the Centre for Effective Altruism's board." https://funds.effectivealtruism.org/funds/ea-community#you-have-concerns-about-conflicts-of-interest-or-grantmaker-independence
MacArthur Foundation COI Policy: Link and selected sections below
An entity includes a corporation, partnership, limited liability company, trust, organization, coalition, commission, university or institute (including a school, department, center, committee, or research project within a university or institute).
The principal executive officer includes the executive head or co-head of an entity, including the principal investigator of a research project or the co-chair of a commission or other entity.
A material affiliation with an entity or individual exists when a director, staff, an investment committee member, or a related party has any of the following types of relationships with the entity or individual:
A material financial interest with an entity exists when a Foundation director, staff, an investment committee member or a related party:
A grant is material to an entity when the amount of the grant is in excess of five percent (5%) of the revenue of the entity.
Whether a director, staff, an investment committee member or a related party derives a “significant personal benefit” or has a “relationship to provide goods or services that is significant” will depend on the facts and circumstances of each case, including an assessment of whether an objective person would consider the benefit capable of affecting the individual’s objectivity or independence.
This Policy is intended to cover any proposed grant, investment or other Foundation business transaction in which there is a conflict of interest.
A conflict of interest will be present if an individual knows that he/she or a related party has a material affiliation with or a material financial interest in the entity or with the individual involved in the transaction, or will otherwise benefit financially or derive a significant personal benefit as a result of the transaction.
The Foundation will not proceed with the following transactions in which a conflict of interest is present:
In all other transactions involving a conflict of interest, a disinterested decision-maker (the board, the investment committee, or the president, as the case may be) will determine whether proceeding with the transaction is in the best interests of the Foundation after considering all the facts and circumstances.
PARTICIPATION IN PROCESS
Directors, investment committee members, and staff who have a conflict of interest regarding a proposed grant or transaction should not vote on or approve the grant or transaction. In addition, a staff member should not work on the grant or transaction where the conflict is present and, unless asked by another director or the president, a director, an investment committee member or a staff member should not participate in formal or informal discussions of such grant or transaction.
The chair of the board or the chair of the investment committee, as the case may be, in consultation with the president and the general counsel, will determine whether a director or staff member with a material affiliation should also be excused from the meeting when the matter is being discussed.
The prohibition against participation in the grant process does not apply when an individual is affiliated with a grantee at the written request of the Foundation. Any such request must be made by the president in the case of a staff member and approved by the chair of the board in the case of the president.
RECORD AND REPORTING
The Foundation will maintain a record of actions taken when there is a conflict of interest present with respect to any grant or transaction.
The general counsel will provide an annual report to the audit committee reflecting all transactions in which there was a conflict of interest and the actions taken.
Bill and Melinda Gates COI Policy: Link and selected sections below
5. OFFICER/BOARD POSITIONS:
Q. I am (or a family member is) an officer/board member of an organization that has received, or might receive, a grant from the foundation. Is this a conflict of interest?
A. A potential conflict of interest arises if (i) you have decision-making authority at the foundation over whether a grant is made to an organization on whose board you serve (or a member of your family serves), (ii) you have (or a family member has) decisionmaking authority over how such funds are expended by the potential grantee and you manage the grant for the foundation, or (iii) you receive (or a family member receives) compensation from the organization because of your position as an officer or director of the organization. • You should disclose your position with the organization to the foundation as required by our Conflict of Interest Policy and also follow the requirements of the other organization’s Conflict of Interest Policy. • You should refrain from exercising decision-making authority with respect to the grant (both at the foundation and at the grantee) and also from assisting the grantee or potential grantee with preparing or submitting grant proposals to the foundation. • In some cases where your family member is directly involved in managing the foundation’s grant at the grantee organization, it also may be necessary to recuse yourself from managing the grant, even if someone else at the foundation approved the grant. If this is the case, please consult the Legal Team.
Q. I am on an advisory board or a nonvoting board member of an actual or potential foundation grantee. Do I have to refrain from making decisions regarding a grant from the foundation?
A. If the advisory board’s recommendations are not binding on the grantee’s board of directors or you do not actually have the right to vote on decisions made by the organization’s board, then it is unlikely that your relationship with the grantee creates an actual conflict of interest. • Nevertheless, you should be sensitive to whether there are any issues of appearance of conflict and not otherwise use your position to exercise undue influence (e.g., if the organization defers to you because you are its tie to substantial foundation grant funds). Your relationship with the other organization should be disclosed to the foundation and you should follow the requirements of the other organization’s Conflict of Interest Policy.
Q. I have been asked to serve as a board member of a foundation-sponsored organization as part of the performance of my employment responsibilities. Is there a conflict of interest?
A. This situation presents one of the most challenging sets of conflicts issues for the foundation. A conflict of interest can exist even though you serve on the board of a foundation-sponsored organization as part of your employment responsibilities at the foundation. • You should disclose your board position with the organization on the foundation’s Conflicts Questionnaire as required by our Conflict of Interest Policy and also follow the requirements of the other organization’s Conflict of Interest Policy. • For the protection of the foundation employee, the foundation and the grantee, and as a matter of good corporate practice, board minutes should reflect a clear record of disclosure and explicit recusal (abstention from voting) by the foundation employee whenever matters on the agenda or under discussion present a conflict of interest between the other organization and the foundation. • A foundation employee serving in this capacity should consciously think about his/her dual roles and which capacity he/she is acting when presented with decisions regarding foundation grants. As a foundation employee, you have certain duties to the foundation, including the obligation to refrain from disclosing, to a grantee or others, foundation confidential information learned in the course of your employment. As a board member, you owe fiduciary duties of care, loyalty and confidentiality to the grantee, which under law may supersede your duties to the foundation as an employee.
o The duty of care requires a director to act in accordance with the best interests of the organization on whose board the director serves, irrespective of other entities with which the director is affiliated, or to which the director owes his or her board appointment. This duty calls upon a director to be informed and to exercise independent judgment when participating in the board’s decisions and its oversight of the organization and its management.
o The duty of loyalty primarily relates to conflicts of interest, corporate opportunity and confidentiality. The duty of loyalty requires that a director (i) be conscious of any potential conflicts of interest he or she may have, and (ii) act with candor and care in dealing with situations in which a conflict exists.
o Confidentiality obligations require that information learned in the course of serving as a director be held in confidence and that board members refrain from disclosing information they receive in this capacity to the foundation or to others without the consent of the organization. You should check with the grantee or contractor regarding its confidentiality policy, and, as appropriate, refrain from disclosing confidential information to the foundation or to others, without prior approval.
o It is important that these fiduciary duties be kept in mind even where you serve on the board of a foundation grantee primarily to ensure good stewardship of foundation grant funds and the accomplishment of foundation programmatic objectives. For example, if a foundation employee serving on the board of directors of a grantee becomes aware as a result of such position (versus through his/her role as a program officer) that the grantee is considering changing the manager of a project funded with foundation funds, it could be a breach of the individual’s duties of loyalty and confidentiality to the grantee to inform the foundation of such change, without the grantee’s prior consent.
• If it is not feasible for you to recuse yourself from participating as a board member of the grantee when decisions with respect to the foundation’s grant to the organization are being made, you should document that your manager at the foundation has approved any final decisions with respect to the foundation’s grant to the organization. The key is that you should not be in a decision-making position on both sides of the grant.
• Finally, the “safe harbors” under Federal tax laws that protect the foundation from being considered responsible for “tipping” a public charity include as a condition the requirement that the foundation not be “controlled directly or indirectly by the grantor.” A grantee is “controlled” for purposes of these rules if by “aggregating votes or positions of authority, the foundation may require the grantee to perform any act which significantly affects its operations or may prevent the grantee from performing such act.” Holding a minority of the grantee’s board seats should typically not satisfy this test, however, where the foundation is the primary funder of the organization, sensitivity to this issue is important.
Q. I am on the board of an organization (which may include a for-profit company) that does not conduct business with the foundation. Do I need to disclose the relationship as a potential conflict?
A. The foundation’s Conflict of Interest Policy requires that employees disclose all of their board affiliations, even if there is no current conflict of interest. • It is important for the foundation to have this information as the foundation might engage with the organization in the future and we need a complete database to identify potential conflicts of interest as part of conducting due diligence. • Note that time spent on board activity on behalf of for profit entities or otherwise not related to foundation business and our mission should be taken as personal, vacation days, and any travel and other related expenses should be paid for personally and not charged to the foundation.
6. FINANCIAL INTEREST:
Q. I have (or a family member has) a financial interest, through ownership or investment, in an entity with which the foundation proposes to do business or to which it may make a grant. Does a conflict of interest exist?
A. Holding a financial interest in an organization does not necessarily create a conflict of interest. It will depend upon the facts and your role as a foundation employee in selecting the entity for the proposed transaction.
• If the employee with the financial interest in the organization does not have a role in selecting or negotiating the terms of engagement or grant with the organization and an independent foundation manger determines that the transaction is fair, reasonable and in the best interests of the foundation, the foundation may proceed with the transaction in accordance with the steps outlined in Question 3 above. Potential appearance of conflict of interest issues should be taken into account in making this decision.
• In all cases, transactions with a for-profit organization in which a foundation employee has a financial interest will be scrutinized closely to ensure that no improper benefits will be derived by the employee as a result of the transaction.
This actually goes back further, to OpenPhil funding CEA in 2017, with Nick Beckstead as the grant investigator whilst simultaneously being a Trustee of CEA (note that the history of this is now somewhat obscured, given that he later stepped down, but then stepped back up in 2021). The CoI has never been acknowledged or addressed as far as I know. I was surprised that no one seemed to have noticed this (at least publicly), so I (eventually) raised it with Max Dalton (Executive Director of CEA) in March 2021 - at least I anonymously sent a message to his Admonymous. In hindsight, it might've been better to publicly post (e.g. to the EA Forum), but I was concerned about EA's reputation being damaged, and possibly lessening the chances of my own org getting funding (perhaps I was a victim of/too in sway to Ra?). Even now part of me is recognising that this could be seen as "kicking people when they are down", or a betrayal, or mark me out as a troublemaker, and is causing me to pause [I've sat with this comment for hours; if you're reading it, I must've finally pressed "submit"]. Then again, perhaps now is the right time to be airing concerns, lest they never be aired and improvements never made. This is what I sent to Max:
Thanks for sharing this, especially since you expressed concerned about speaking publicly.
I have expressed similar concerns within my local EA group because I feel like there is no transparency regarding the CEA, how their money is spent, and the cost-effectiveness of EAG. I was typically met with the reply that EAG was justified because of the networking opportunities and that it being a lavish event was only an "optics" issue. This was especially strange to me since I was new to EA, and it didn't seem to align with what I imagined the community to be (specifically the lack of rigour in how they were deciding to spend money on such an event, or other things they spend money on such as university groups going on retreats). Thanks for making me feel validated in these concerns.
There was a post Linch did earlier this year about their experience as a junior grantmaker in EA. It had an interesting part about conflicts of interest:
It makes sense to me what they were saying, even though it's troubling, because conflicts of interest are a problem.
It seems natural that more due diligence and organizational safeguards (like avoiding CoIs) will be some of the lessons learned from the FTX disaster. This could be an opportunity to prioritize such things more highly to improve EA's worst-case preparedness (e.g. corruption) , even at the expense of some impact/efficiency gains from having the best experts everywhere when things are going well.
I added a note for people to check this comment out.
I think this reflects two facts:
I wouldn't conclude from this that EAs are uninfluenced by our philosophical commitments, in real life. I expect that SBF was influenced in various important ways, though personality, character, etc. surely played a large role too.
Yes, this. Most fraudsters don't have such strongly held views on why the Kelly criterion for determining optimal bet size doesn't apply to them! (SBF did a famous thread on this and Caroline's tumblr has that line about how real EAs endorse high leverage and double-or-nothing flips).
I think it would be wrong to blame utilitarianism per se for what happened because the vast majority of utilitarians absolutely do care about the risk of ruin - as they should - but I think SBF's own brand of EA-aligned thinking (I assume short AI/bioengineered pandemic timelines factored in here) played a huge role in why he took such insane risks.
Did you mean "low moral expectations" instead of "low moral standards?"
I mean both.
From the article:
Why isn't the former FTX Future Fund team doing this too? Instead of the symbolic gesture of resigning, why are they not doing their best to raise other funds in order to keep the funding promises they made? Don't they believe in their own orgs grant evaluations? Because if they do, they should be talking to OpenPhil about it right now.
Or are they already doing this? But then why not tell us? This is not the time for secrecy.
I think they're doing their best to help. From the resignation letter you linked (emphasis added):
Thanks for pointing this out.
I did read the post but obviously missed this part. I apologize.