Effective Altruism Is a Dangerous Cult. Here's Why.
Critics like myself have been the target of a coordinated campaign of harassment, stalking, online impersonators, and threats of physical violence
[Several updates have been added to this document, below. May 4, 2024.]
Note for journalists: I am willing to, on or off the record, corroborate everything I say below that isn’t hyperlinked. Contact me here.
Table of contents:
Introduction
Background
The Censoriousness of Effective Altruism
Deception and Manipulation from EAs
Discussion of the Use of Violence
Harassment, Threats, and a Defamatory Article
An Ongoing Campaign of Harassment and Threats
The “Mark Fuentes” Article Didn’t Come Out of Nowhere
The Origins of “Mark Fuentes”
Appendix: Misrepresentations by Andreas Kirsch
1. Introduction
The Effective Altruist (EA) and longtermist communities have become extremely powerful over the past decade. They are backed by billions of dollars, are shaping the policies of the United Nations and UK, and have fostered connections with some of the wealthiest people on Earth, such as Elon Musk. Many EAs and longtermists believe they are literally saving the world, and longtermists in particular are motivated by a techno-utopian vision of the future in which we acquire “posthuman” capacities like immortality, colonize the accessible universe, and generate “astronomical” amounts of value by creating huge computer simulations full of digital people. As one EA writes: “So you want to save the world. As it turns out, the world cannot be saved by caped crusaders with great strength and the power of flight. No, the world must be saved by mathematicians, computer scientists, and philosophers”—i.e., them.
These communities are also highly centralized and—despite their own claims otherwise—very intolerant of critics. In particular, many EAs welcome what could be called “soft” criticisms that focus on this or that component within the general EA framework. In some cases, such criticisms are encouraged. What the community strongly dislikes are critiques that target the framework itself—for example, by arguing that longtermism could be extremely dangerous because it contains the very same core ingredients (i.e., utopianism plus a broadly utilitarian mode of moral reasoning) that have motivated apocalypltic movements of the past to engage in violent, even genocidal, acts. Or by arguing that longtermism is just another iteration of what scholars call the “eternal return of eugenics,” which I fully believe to be the case.
Those who launch such critiques are typically dismissed out-of-hand as being “bad-faith” actors, which is often false. Worse, the EA and longtermist communities have a long track record of trying to silence, intimidate, bully, threaten, and otherwise “cancel” their critics. The EA leadership has shown that they are willing to bend the truth when it suits them (e.g., about Sam Bankman-Fried), and many community members are so afraid to voice criticisms of fundamental EA dogmas that they either self-censor or publish under anonymous or pseudonymous accounts. Some EAs are known to have created anonymous online accounts for “deceptive” purposes, and numerous people from the community have told me that talk of using violence to achieve their ends sometimes comes up. Just 12 months ago, I would have been hesitant to describe EA and longtermism as a “cult.” Today, I wouldn’t hesitate. In many significant ways, the EA movement resembles Scientology, and in fact people I follow on Twitter have started to call longtermism—one of the main “cause areas” of EA—the “Scientology of Silicon Valley.”
This article surveys the evidence supporting these claims: that EA is a cult that is intolerant of critics. Part Two examines some examples of how EAs have tried to silence critics, and how the community has created a culture of intolerance that stifles the free exchange of ideas. We will then examine some instances of duplicity among leading EAs, including Holden Karnofsky, as well as some reasons that critics of EA might be warranted in worrying about their physical safety. Much of this will serve as background for the main focus of this article, in Part Three: the coordinated and relentless campaign of harassment that has targeted me since I began to regularly publish critiques of EA and longtermism last summer. This campaign has included threats of physical violence, online accounts impersonating me, anonymous emails threatening to dox me and inappropriately referencing my ex-wife, and a barrage of anonymous social media accounts that have harassed and stalked me and my colleagues, often hurling profanity-laced insults and spreading demonstrably false lies about me. The final subsections of Part Three will look at the origins of a defamatory hit-piece written by “Mark Fuentes” (an earlier draft of which was published by “Claressa Meals,” whose material came from the EA Forum), and then some misrepresentations on Twitter by Oxford DPhil student Andreas Kirsch, in the appendix. The hit-piece perfectly illustrates how, when the EA community has nothing of substance to say to critics, its members will resort to underhanded and dishonest tactics to intimidate their perceived enemies. I can assure readers that the article will get more shocking as it progresses.
As a result of the harassment detailed here, I have become genuinely worried about my physical safety, and the campaign of harassment, including the “Mark Fuentes” article, has been emotionally draining. Nonetheless, all of this has convinced me, even more than before, that EA is a dangerous cult, and that people should be very alarmed by its rapid growth and power.
2. Background
The Censoriousness of Effective Altruism
Leading EAs will take measures, sometimes drastic, to “cancel” critics. To quote David Pearce, who cofounded the World Transhumanist Society with Nick Bostrom and considers himself an EA: “Sadly, [Émile] Torres is correct to speak of EAs who have been ‘intimidated, silenced, or ‘canceled.’” Here are some examples, beginning with my own experiences:
In August 2022, I published an article in Salon about how “toxic” the EA and longtermist communities are. As if to prove part of my point, members of the EA community based in Oxford emailed my editor to complain. They spent more than a week trying to convince him to make extensive revisions or retractions, as well as to not publish me in the future (i.e., to fire me). One, a spokesperson for a prominent figure of the movement, insisted on an “informational” Zoom meeting with my editor, seeking to make the case that my article was unfair and poorly substantiated. My editor did not agree. In emailing back and forth with Salon about the incident, the term “MacAskill-industrial complex,” a reference to William MacAskill, was bandied about. (This was not my coinage, by the way.)
Something similar happened after I published another article criticizing longtermism: a prominent member of the community inundated multiple people involved for more than two weeks with demands. I am not at liberty to say more, but investigative journalists are welcome to contact me for details. It was intense and extreme.
The pressure to confirm to orthodoxy within the community is significant. After publishing my 2021 article on longtermism in Aeon, a colleague of mine had to withdraw from a collaboration we were working on due to pressure from within the community. People have been explicitly told to never work with me, simply for criticizing the longtermist ideology. Even earlier, in 2019, when I considered myself to be a longtermist, a prominent figure in the community told me over Skype that if I didn’t stop publishing political articles for outlets like Salon, I would lose funding and collaboration opportunities in the future. (Many EAs strive to be “apolitical,” based on the belief that, in Eliezer Yudkowsky’s words, “politics is the mind-killer.”) This isn’t so much an instance of censorship, but it did strike me as overly controlling. I did not believe it was their place to tell me what I can and cannot write about, and in fact this incident is what convinced me to start speaking out publicly about the problems I saw with longtermism, EA, and their respective (largely overlapping) communities. I experienced something similar after publishing an article, also in 2019, that was critical of Steven Pinker, someone EAs had been wooing for some time. Numerous EAs were upset that I hadn’t been given “permission” from MacAskill to publish the article in Salon, which once again struck me as overbearing and weirdly authoritarian. These incidents raised red flags.
I am not the only one with such experiences. In late 2021, Carla Cremer, a former EA, published an article on the Effective Altruism Forum (EA Forum) describing her experience trying to write a critique of longtermism with Luke Kemp. It sounds rather traumatic. Here’s what Cremer wrote:
She adds that EAs who disagreed with her and Kemp accused them “of lacking academic rigour and harbouring bad intentions.” She continues, noting that some people actively attempted to prevent the paper’s publication:
(“TUA” stands for “techno-utopian approach,” the dominant paradigm within longtermism.)
Just last month, ten EAs published a lengthy critique along with suggestions of how to reform the EA movement, such as that EA should become more “democratic.” Of note is that all ten authors felt the need to conceal their personal identities, choosing to publish instead under the name “ConcernedEAs.” Why do this anonymously? Because they are afraid of retribution from EA—of getting fired, being denied funding, and being ostracized from the community. Here’s what they said:
This is worrisome. The EA community claims to be open to criticisms. As Ezra Klein describes the movement, “it’s always, always questioning its own assumptions and everyone else’s.” In truth, EA is so intolerant that its own members don’t feel safe voicing criticisms under their own names. Two people familiar with the “ConcernedEAs” group even contacted me prior to the article’s publication asking me not to tweet anything about it for a week or two after it’s posted. Why? Because even the most tenuous connection with a critic like myself could lead other EAs in the community to dismiss the “ConcernedEAs” article out of hand. This is how allergic the community is to having its framework critiqued. As Dr. Sarah Taber, an ex-Mormon, wrote on Twitter about this article: “Reading through it, all I can think is ‘Ohhhh yeah. I remember this stage of trying to break out of a cult.’”
I have also heard from many EAs over the years—dozens, at least—who tell me, in private, that they agree with my criticisms and appreciate my work, but are too scared to say so publicly for fear of losing funding, being ostracized, and so on. They deliberately choose to “self-censor.” I have also been contacted by family members of EAs, including some high-level EAs (!), with concerns that they have joined a cult. Once again, journalists are welcome to ask me for details.
Deception and Manipulation from Notable EAs
In addition to the censoriousness of EA, another issue that will become relevant below is the fact that (a) several prominent EAs have been caught using anonymous accounts online for purposes of deception and manipulation, and (b) the community’s leadership has a history of bending the truth to promote the EA brand.
Taking these in order, let’s begin with this tweet from Kerry Vaughan, which was a response to me reporting that I’d received an anonymous email threatening to dox me, one day after I pushed a short thread on social media about how large portions of William MacAskill’s book weren’t actually written by him. Vaughan wrote:
I encourage people to read the thread in its entirety. It goes into detail about the “targeted harassment” that Ryan Carey perpetrated, which he admitted to. Most troubling, though, was this:
As noted, Carey founded the EA Forum, and Vaughan notes that he also “worked at MIRI, Open AI, FHI, and DeepMind.”
Another alarming example involves Holden Karnofsky, cofounder of Open Philanthropy and GiveWell, and someone who William MacAskill reports “had a particularly profound impact on my broader thinking about longtermism.” (OpenPhil and GiveWell are the two biggest funders of EA.) GiveWell conducted an investigation that was first published (I believe) in 2011. Under a section titled “Anonymous and deceptive online promotion,” the report states:
Holden Karnofsky:
- Posted a question to Ask Metafilter using the name “geremiah” and asking for help figuring out where to donate; answered the question using the name “Holden0” with a link to GiveWell.net; and posted a followup comment under the name “geremiah” praising GiveWell and designating Holden0’s response as the best one. This can be seen here.
- Posted two comments to Lifehacker under the name “geremiah”; the first asked for help finding a good charity, and the second linked to GiveWell. The first was deleted by Lifehacker; the second can be seen here.
- Sent ten emails to bloggers recommending that they blog about GiveWell, using the tone of an unconnected party and using the name of staffer Teel Lidow.
- Commented on Wall Street Journal online using the name “Research24” and linking to GiveWell. (Note that Holden has no record or recollection of making this comment, but recognizes the language and account as his own; this comment was pointed out by a commenter on The GiveWell Blog, after the Board had already met and addressed misbehavior in this general category.)
This shows that some high-level EAs have used anonymous accounts to dox and deceive people for the purposes of promoting EA and attacking what EAs see as threats.
Turning now to the issue of bending the truth, consider the case of Sam Bankman-Fried, one of the most prominent EA longtermists, who now faces 115 years in prison for organizing “one of the biggest financial frauds in American history.” (Bankman-Fried wasn’t the first crypto billionaire involved in criminal activities that EA has embraced, by the way: Ben Delo was another.) I will let Gideon Lewis-Kraus of The New Yorker explain:
The story commonly told about Bankman-Fried was that he drove a beat-up Toyota Corolla, slept on a beanbag, and had nine roommates. MacAskill repeated this fable to me, characterizing it as evidence of Bankman-Fried’s profound commitment to the cause. What he did not mention, and what came out only in the last few weeks, is that Bankman-Fried and his roommates were living in a forty-million-dollar penthouse in a gated community in the Bahamas—part of a total local property portfolio worth an estimated three hundred million dollars. His parents, professors at Stanford Law, owned a vacation condominium worth millions of dollars.
Lewis-Kraus continues:
E.A. leadership ratified a mythology about Bankman-Fried that was simply not the case. One senior member of the community told me that the peculiar contradictions of Bankman-Fried’s life style were widely known but somehow unexamined: it was true that he drove a beat-up Corolla, but it was also true, if underemphasized, that he enjoyed a sumptuary existence—not only the lavish penthouse but the use of such appurtenances as a private jet.
All cults have their charismatic leaders, and frequently the image of these leaders is built on falsehoods and exaggerations propagated by members of the community. EA is no different: the community has demonstrated a willingness to lie (in this case, by omission) to prop-up its leaders and promote its brand. The public perception of EA is something the community cares deeply about, which I document in this article for Salon. As Simon Knutsson, who’s spent time in and around the EA community, told me for an article I published two years ago in Current Affairs:
Discussion of the Use of Violence
So far we’ve covered attempts by EAs to convince my editors to stop publishing me, worries from EAs that criticizing the ideology could have serious professional consequences, examples of leading EAs concealing their identities for purposes of deception and manipulation, and a general disregard for the truth when doing so suits the community. Let’s now examine some even more worrisome issues.
Several people since last summer have privately contacted me to say that it wouldn’t be irrational for me to worry about my personal safety, given the reputational damage I may have caused the EA movement with my articles. This is especially the case if I were to visit Oxford or the San Francisco Bay Area.
In fact, Knutsson himself wrote in 2019 that “I am concerned about my safety” because the articles he’s published could undermine the longtermist project of creating a techno-utopian world among the stars full of astronomical amounts of “value.” As I alluded to at the beginning of this article, longtermism is very much influenced by utilitarianism, and history is overflowing with movements that, by combining a utopian vision of the future with a broadly utilitarian mode of moral reasoning, came to believe that extreme, violent, even genocidal actions may be “justifiable” for the greater good. As Pinker writes in The Better Angels of Our Nature:
This is, in part, why I am genuinely worried that “true believers” of longtermism could find themselves in a situation where violence toward certain people seems, to them, necessary and warranted, given the astronomical stakes. Others have expressed the same concern, including Peter Singer and Olle Häggström (the latter of whom otherwise speaks favorably of longtermism). Knutsson himself foregrounds this aspect of longtermism, writing that he’s
most concerned about someone who finds it extremely important that there will be vast amounts of positive value in the future and who believes I stand in the way of that. … [A]mong some in EA and existential risk circles, my impression is that there is an unusual tendency to think that killing and violence can be morally right in various situations, and the people I have met and the statements I have seen in these circles appearing to be reasons for concern are more of a principled, dedicated, goal-oriented, chilling, analytical kind.
He adds that “if I would do even more serious background research and start acting like some investigative journalist, that would perhaps increase the risk more. Simply one more reason to try to make me stop and be quiet, and I feel concerned enough about my safety as it is” (italics added).
I do not believe that Knutsson is being paranoid, nor do I believe that worrying about my own safety is unwarranted. Indeed, I have assumed the role of an investigative journalist for some of my critiques of EA and longtermism, as when I discovered an old email from Nick Bostrom, a hugely influential philosopher among EAs, in which he wrote that “Blacks are more stupid than whites” and then mentioned the N-word. Since I stand in the way of utopia by publishing critiques of EA and longtermism, some of which have gained lots of attention, it is not far-fetched to imagine a “true believer” deciding that it would be best if I no longer existed. To underline the seriousness of this situation, a former leader in the EA/Rationalist community contacted me last year and told me that the Bay Area community has become, in their words, a “full grown apocalypse cult.”
Even more alarming, I recently published a Twitter thread detailing some of the harassment and threats I’ve received from anonymous EAs over the past few months. In response, someone contacted me about a 3-day-long workshop they attended in Berkeley, where longtermist organizations like the Machine Intelligence Research Institute (MIRI) are based, on how to create “safe” artificial general intelligence (AGI).1 (Many longtermists believe that AGI will either annihilate humanity or usher in a utopia.) During the meeting, some participants explicitly considered the possibility of targeted killings, including by sending bombs through the mail. Here is a screenshot from the meeting minutes that was shared with me:
Another section of the document said this:
The reason this person contacted me was because, to quote a message they sent to someone else, “when I saw that Torres was threatened with being kneecapped I took it as real. I wanted to warn them of my own concerns” (reproduced with permission). Although the discussion above focused specifically on AI researchers, the proposal to “be ted kaczynski” could easily be redirected at critics and investigative journalists like myself. These ideas are out there, in corners of the community. How long before a true believer acts on them?
Update: Consistent with my fear, going back to my 2021 Aeon article, that “longtermism” could be used to justify extreme violence, Eliezer Yudkowsky published an article in TIME magazine arguing that we should risk thermonuclear war to avoid the AGI apocalypse. When he was later asked “How many people are allowed to die to prevent AGI,” he had the following jaw-dopping response:
This community is dangerous. If someone with political power in a nuclear nation were to take Yudkowsky seriously, the result could be an unthinkable catastrophe.
3. Harassment, Threats, and a Defamatory Article
An Ongoing Campaign of Harassment and Threats
This leads us to the campaign of harassment and threats to my physical safety that I have been dealing with from the EA community since last summer. The present subsection will survey a few examples, and the next will turn to a defamatory article published last November by someone with the pseudonym “Mark Fuentes.” As always, I welcome conversations with journalists interested in pursuing this story further.
Virtually all of the harassment and threats that I’ve received have happened immediately after either (a) the publication of critiques of EA/longtermism in outlets like Salon and Truthdig, or (b) tweets or Twitter threads that I posted about these ideologies and their corresponding communities. In the majority of cases, incidents have involved anonymous accounts explicitly upset with my articles about EA and longtermism. These accounts are almost always created around my publications, and almost all of their posts are about me and my articles.
Consider the following DM that I received on October 11, 2022, which was two days after I criticized a prominent EA longtermist for claiming to be “generally quite optimistic” about climate change, and three days after I tied longtermism to eugenics, which has angered many longtermists (even though, as I show in detail, it’s accurate). The message was this:
Around the same time, another anonymous account was created. As of this writing, it is still online with the same profile impersonating me. Not only does it include the trans flag next to my name (at the time, my account included a trans flag), but lists my personal website and locates me in Germany, which is where I currently live. As you can see, the profile description is “Kneecap breaker.”
On October 12, one day after the anonymous threat above, this account posted its first tweet, which said:
Nearly every tweet from this account is in response to tweets that I have posted, including statements like “I wish you remained unborn” and, under a picture I posted of myself, “Dam u ugly ash” (followed by two laughing faces). I take it that “ash” means “as shit.”
Later that month, just two days after my Bulletin article on longtermism was posted, I received the following anonymous email from a Guerrilla Mail account. I found the line “before it’s too late, buddy” to be vaguely threatening:
This was, in fact, one of three emails I’ve received from a Guerrilla Mail account. Another made an inappropriate reference to my ex-wife and called me a “psycho,” while yet another threatened to dox me by promulgating personal information that I once shared among friends on Facebook, and which would be wholly unacceptable to spread without my permission. These were sent two days apart, on October 20 and 21, after I had posted a short thread criticizing William MacAskill’s book What We Owe the Future—as noted above in my discussion of Ryan Carey.
Yet another anonymous account on Twitter, created in October 2022—just like the account above—inappropriately tagged my ex-wife on October 19 under a thread I’d posted about coming out as nonbinary. The thread did not explicitly mention my ex-wife.
Another account created in October 2022, going by the name “Claressa Meals” was reported by online friends for harassment. She was inundating followers of mine with an article she’d posted on Medium that we will discuss more below.
In every case, above and below, these accounts were created after I started criticizing EA and longtermism last summer. Most were created in September and October, although a few others were created last December. Their sole purpose has been to harass me and my colleagues on Twitter. Some of the accounts have been deleted, while others have gone quiet.
Consider, for example, yet another account created in October of last year with the name “Mark Fuentes.” On my count, they posted 138 tweets or retweets between November 8 and 25, a period of seventeen days. Of these, a total of 123 were explicitly about me. That’s an average of 7.2 tweets about me per day.
Another account also created in October under the name “Howard Hunter” has posted literally hundreds and hundreds of tweets about me since just last November, most of which share a defamatory article by “Mark Fuentes” (below) in response to people who’ve posted favorably about my articles on EA and longtermism. In some of these tweets, the acocunt calls me a “liar.” They refer to my Aeon article as a “complete fabrication.” They have claimed that I “spread conspiracies.” In one response to someone he wrote: “What you don’t seem to understand is that Torres is full of shit. It doesn’t matter if he attacks or defends. Now you know the truth. Deal with it.” On my count, he tweeted a total of 103 times between just January 15 and February 5, 2021. Ninety-nine were explicitly about me, and 75 shared the “Mark Fuentes” article. That’s an average of nearly 5 tweets about me per day, although most of his tweets have been sent in clusters. This 3-minute-long video shows a fraction of the hundreds of tweets he’s posted under people sharing my articles. This is disturbingly obsessive behavior, in my opinion:
Yet another account was created last December, which describes itself as a “parody.” It’s profile picture looks like my profile picture run through an AI program to make me look older (in my profile picture, I’m wearing a hat and sitting in front of a whiteboard with writing on it):
The sole purpose of this account is to mock me. In one tweet referencing a request I posted on Twitter to help find a defamation lawyer (because of the “Mark Fuentes” article), they wrote: “Does anyone know of any affordable defamation/libel lawyers in the UK? Someone’s been accusing me of calling people nazis!! (When really I just called them ‘literally Hitler’!).” Another tweet seems to reference a false claim that numerous anonymous accounts defending EA/longtermism have made about me, namely, that I favor sterilizing the poor. These tweets are often made in reference to my claim that longtermism is rooted in eugenics, in a clumsy attempt to show that I’m a hypocrite. This is one of the more truly bizarre accusations I’ve come across.
This is a brief snapshot of the threats and harassment that I’ve received over the past seven months. I have documented many more incidents than this, but for the purposes of this relatively short article, listing them would be tedious. Journalists are welcome to contact me for more details.
Update: on June 26, 2023, I received an anonymous email from someone in the EA community referencing a short film titled “This Is Vanity,” about the murder-suicide of a mother who kills herself and her daughter by blowing up a car while both are inside. The email states: “I hope it will take something far less extreme than what happens in the film to make you look at the kind of person you're becoming.” This was clearly written by someone in the EA (I’d now say “TESCREAL”) community because it closes with this line: “It is heartbreaking to see people who once shared similar goals descend into infighting, but hopefully we can look back on this as just a phase that we grew out of.” The words “once shared similar goals” is a clear reference to the fact that I used to be a TESCREAList myself. I contacted friends of mine with some expertise on this sort of thing, including a lawyer, and all strongly recommended that I report this to the authorities. End up update.
Clearly, many EAs are obsessing over me and my critiques. Some of these accounts have been created for the sole purpose of harassing me and my colleagues, friends, and acquaintances, often utilizing ad hominem attacks and spreading falsehoods about my person. All of this is in addition to the concerted efforts by EAs at Oxford to, essentially, get me fired from Salon and elsewhere by flooding my editor’s inbox with demands and complaints.
When the flurry of ongoing threats and harassment that I’ve received over the last half-year are put in the context of the previous section, the situation looks very worrisome indeed: EAs and longtermists literally believe they’re saving the world; leading members of the community are known to have created anonymous accounts to harass, dox, and deceive people; there is talk among some in the space of EA about using violence to achieve their ends; and I am precisely the sort of individual—a critic and journalist—that Knutsson notes would be targeted by this techno-utopian cult. Numerous people have told me that I should be somewhat concerned about my safety, and I am anxious about what might come next. (Will I receive more threats for publishing this article?) It was clearly wise for the ten “ConcernedEAs” to have published their critique of EA without revealing their true identities.
Finally, let’s turn to the most recent example of harassment: a defamatory article by the aforementioned “Mark Fuentes.”
The “Mark Fuentes” Article Didn’t Come Out of Nowhere
Last November, “Mark Fuentes” published an article that manipulates and rearranges screenshots from me and people I’ve interacted with, while cutting out the social media posts I was responding to, along with those supporting my claims, to patch together a heavily edited portrayal of me that does not correspond to reality. Many of these screenshots are from late 2017 or early 2018, and many involve my interactions with people like Helen Pluckrose, James Lindsay, and Peter Boghossian. In a subsequent article, I will show how what actually happened is literally, and consistently, the opposite of what “Mark Fuentes” claims. As noted above, this perfectly illustrates just how low EAs are willing to go to undermine critics: if they don’t have anything of substance to say about my critiques, they will resort to personal attacks in hopes of intimidating and, ultimately, silencing me.
For now, let’s take a look at where this article came from. While the article is somewhat cleverly put together, it’s also quite sloppy, and “Mark Fuentes” left a trail of digital evidence that overwhelmingly points to a particular EA-longtermist as it’s author. I will not mention this person’s name here, but rather refer to them as “X.” (As it happens, “X” was at the center of one recent incident of harassment—I don’t know what else to call it—involving some people I work with; I would be happy to provide details to investigative journalists or confidants curious about this.) “X” has also been obsessed—again, I don’t know how else to put this—with me since 2019, and has frequently spread the exact same false claims about me that one finds in the “Mark Fuentes” article. Not only is there evidence of this all over publicly accessible websites, and not only was this person involved in the harassment incident alluded to above, but I have received numerous messages over the past seven months from people reporting their own unsettling experiences with “X’s” obsession with me. Here is just one example: an unsolicited message from someone I didn’t know or follow on social media at the time. They wrote:
When I asked someone in the longtermist community about this person, they responded with this:
I am genuinely frightened by “X.” They scare me, and I am worried about the situation escalating. I am worried about my physical safety. In what follows, I will provide some reasons for believing that this person is behind the “Mark Fuentes” article, although this will be a small fraction of the evidence that I have gathered in a 29-page (and growing) document. Journalists interested in reporting on harassment and abuses of power in the EA and longtermist communities, please contact me. I would love to share this document with you.
The Origins of “Mark Fuentes”
The article by “Mark Fuentes” didn’t come out of nowhere. An earlier draft was posted on Medium on October 15, about three weeks before the “Mark Fuentes” article went up, written by someone going by “Claressa Meals.” It was titled “Emile Torres’ harassment,” which is somewhat ironic given that, as noted above, online acquaintances of mine were compelled to report “her” for harassing us. (Note: sometimes the Wayback Machine archive of the webpage doesn’t load properly, sometimes it does. However, I have a screengrab of the entire thing if you’d like to see it.) Both the “Mark Fuentes” and “Claressa Meals” articles make the exact same false accusations, and the “Mark Fuentes” article contains many of the very same screenshots as the “Claressa Meals” one.
Where did these screenshots come from? The very first place they were posted online was—and perhaps you could see this coming—the EA Forum. They can be traced back to three accounts: an anonymous account that I will here call “ONE,” another anonymous account that I will here call “TWO,” and a third account belonging to “X.” In 2021, ONE posted a single screenshot on the EA Forum of something Pluckrose said about me. The screenshot was accompanied by information that clearly indicates that it was “X.” This information specified dates and details about their relationship to me that fits only a single person in my past: “X.”
[EDIT: Adding this hear because you won’t find it via a Google search or searching the EA Forum, so it doesn’t give away their identity: ONE wrote on the EA Forum, “I'm writing this from a throwaway account because the last time I criticized Torres in another forum two years ago, under my real name, he started harassing me. Unfortunately, this does not appear to be an isolated incident, as I found when discussing this episode with other people who had also crossed him in public,” a reference to the right-wing troll Helen Pluckrose, whose claim-to-fame is literally deceiving dozens of academics and submitting fabricated data to academic journals. ONE then posted the screenshot that they later deleted, five months before the very same screenshot ended up in Mark Fuentes’ article. From the start, “X” and “ONE” have been claiming that I “harassed” them. This is false, and I have my DMs with “X”—the person who criticized me using his real name exactly two years before ONE’s post!—saved to prove this. The guy has been obsessed with me since hallucinating that I “harassed” him, which I didn’t, and which I can prove. END OF EDIT.]
Then, in 2022, TWO posted an article on the EA Forum that made a false claim about me—a claim that “Mark Fuentes” repeats in his article, and “X” has also repeatedly made. This claim concerned me and two members of the EA community who I once worked with, and those two members have confirmed that this claim is false.
Of note is that under this post by TWO on the EA Forum, both TWO and “X” posted different screenshots of me dating back to early 2018. These are the very screenshots that ended up in the “Claressa Meals” Medium article. Also of note is that the original post by TWO was only up on the EA Forum from September 13 to 15. That’s just three days. So, how did this “Claressa Meals” person manage to find it? They must have been incredibly lucky to have been at the right place at the right time—or “Claressa Meals” is TWO, and this common identity explains how “she” managed to find all the screenshots from TWO and “X” under TWO’s post.
A couple further points of interest: first, “Claressa Meals” presents the very same screenshots as TWO in almost the exact same order. In fact, many of these screenshots are pixel-for-pixel identical. Second, “Claressa Meals” decided to redact certain names, like “Michael Shermer,” from screenshots of things I’d written.2 And so did “X.” That is to say, “Claressa Meals” took the screenshots from TWO and redacted all the same names that “X” had redacted in their screenshot, which “Claressa Meals” also reproduced. Not only did “Claressa Meals” redact the same names as “X,” but “she” redacted them in exactly the same way: using white rectangles with thin black borders. But (a) why redact any names in the first place? How odd. And (b) why redact them in the same idiosyncratic way? It’s as if the same person—“Claressa Meals” and “X”—had redacted all these screenshots on the same computer, using the same image editing software. This is very strange, but it gets stranger.
“Claressa Meals” tried hard to get people on social media to pay attention to their article. Once again, their account was reported to Twitter for harassment, and no longer exists—I am not sure if she deleted it or was banned.
But the article failed to get any attention. The following month, in November, three weeks after the “Claressa Meals” article containing screenshots from the EA Forum was published, “Mark Fuentes” published their article. At the very same time, the “Claressa Meals” article was removed from Medium. If you click on the link, you will get a 401 message saying: “The author deleted this Medium story.”
As noted, the “Mark Fuentes” article contains many of the very same screenshots and false accusations as the “Claressa Meals” article. The content and central theses are identical, although the “Mark Fuentes” article greatly elaborates these, rearranging things to make me look as though I am in the wrong.
But here’s the clincher: the “Mark Fuentes” article also includes the screenshot posted by ONE on the EA Forum back in 2021, which “Claressa Meals” doesn’t include. There are two important points to make about this: first, “Mark Fuentes” doesn’t merely include the same screenshot of what Pluckrose said about me, it’s literally the same screenshot, pixel for pixel. This, as with everything else, I can prove. The significance of this is that ONE deleted their comment from the EA Forum—and along with it, the screenshot—on May 14, 2022, more than five months before “Mark Fuentes” was on the scene. The identical screenshot that “Mark Fuentes” borrows from ONE is not on the EA Forum. It cannot be found via Google, and you cannot search the EA Forum via the Wayback Machine (i.e., the EA Forum search function doesn’t work on the Wayback Machine, so it’s not possible to, say, search my name in the EA Forum through the Wayback Machine). There is just no way to find it unless one has the Wayback Machine URL (which is how I found it: I had archived this comment from ONE and saved the URL).
[EDIT: to be clear, the screenshot that Fuentes included in his article appeared once online prior to his article: on the EA Forum. And this instance of the screenshot was deleted from the EA Forum. For five months prior to Fuentes’ article, there was not a single website that included this screenshot (click the link)—it simply didn’t exist on the Internet. END OF EDIT.]
So, how did “Mark Fuentes” get ahold of it? He claims to be someone who only recently came across me, sometime late last year. The only explanation is that “Mark Fuentes” is ONE. There is no other way for him to have found ONE’s screenshot for his article. And since ONE clearly indicates that they are “X,” and since “X” has been spreading, over the past many years, the very same false accusations about me that one finds in posts by TWO, “Claressa Meals,” and “Mark Fuentes,” there is little doubt about who the culprit is.
In fact, “X” and TWO have made quite literally the same accusations about me on the EA Forum—in some cases verbatim. All of this is documented, once again, and I would be happy to share the evidence with journalists. Here’s a diagrammatic summary of where the “Mark Fuentes” article came from:
Let’s take a step back. The Twitter account of “Mark Fuentes” describes him as “Kind, polite, and stands up to bad online behavior.” But consider the situation: out of nowhere, “Mark Fuentes” suddenly creates a Twitter account last October, when many of the other accounts that harassed, stalked, impersonated, and threatened me were also created. This person then posts the equivalent of a 32-page-long Substack article in early November, at exactly the moment that the “Claressa Meals” article was removed by its author. This “Mark Fuentes” article recapitulates the very same claims as ONE, TWO, “X,” and “Claressa Meals,” and includes the very same screenshots—sometimes identical screenshots—as ONE, TWO, “Claressa Meals,” and “X,” all of which were previously posted on the EA Forum. “Mark Fuentes” has since completely disappeared: his last tweet was December 9.
He claims to care about “bad online behavior.” But does any of this make sense? A random person with a tenuous connection to EA (according to his article) appearing out of thin air, who becomes intimately familiar with stuff I’ve written since 2017, writes a protracted Substack article, posts an average of 7.2 tweets about me per day for two weeks straight, and then vanishes. Remember the meme:
The obvious answer to these riddles is that “Mark Fuentes” isn’t just some guy who’s honestly concerned about “bad online behavior.” Rather, “Mark Fuentes” is an EA who published his article in an attempt to hurt my reputation by defaming me. “Mark Fuentes” has known about me for years, and his article is just another manifestation of the coordinated campaign to intimidate, harass, and frighten me. This is EA. This is longtermism. People in these communities will stoop to this level, because they believe that they are literally saving the world.
I asked “Mark Fuentes” to verify his identity. He claims to be a public defender in New York, and to have attended the funeral of Aaron Swartz. I told him that I was worried that he might be involved in the harassment and threats that I’ve received from EAs, which he denies in his article. He responded that he doesn’t want to do this because I would “harass” him if his identity were revealed. So I formulated the following proposal, which would enable him to maintain his anonymity and reassure me that he is not, in fact, the person who’s been harassing, impersonating, and threatening me. I presented this proposal to him on Twitter and via his Substack page. Here is what I wrote him, just before sending it. I have redacted the handle of TWO in it:
Shortly after, he disappeared and has yet to be heard from, despite several EAs on Twitter agreeing to be a confidential go-between.
I understand the lengths to which this community will go to preserve their image and power. This is why I’ve become scared of EA. This will not prevent me from continuing to publish criticisms of them. To the contrary, it confirms my suspicion that this is a dangerous cult.
Once again, I can prove everything I say above. Everything is documented. Some claims lack hyperlinks because, at this stage of my investigation, I do not want to reveal who these people are. However, as noted, I would be willing to share details with journalists or (in confidence) colleagues and friends.
The EA and longtermists community have been dealt several major reputational blows over the past few months, from the embarrassing collapse of FTX to the ignominious scandal involving Nick Bostrom’s old racist emails. Just recently, Time magazine published a shocking article about how pervasive sexual harassment and assault is within the community. What I have written above could be yet another scandal. I can assure readers, and journalists, that it’s an even bigger story than it might seem, for reasons that would become immediately apparent if I were to divulge them. I am eager for the full story to see the light of day sometime soon.
[EDIT: an anonymous user posted the Fuentes article to the EA Forum on May 1. Someone claiming to be “Mark Fuentes” has been commenting under it. Interestingly, “X” just recently deleted their EA Forum account—right before this anonymous user and “Mark Fuentes” showed up on the EA Forum. My guess is that they worried that people might look into the Fuentes article and find that his points are almost identical to those (sometimes in their wording) made by “X.” By deleting his EA Forum account, “X” makes it much harder to find earlier comments from him discussing me (although I have documented all of these comments, and have screenshots).
The comments thread under the anonymous post also includes the following, which, as BrownHairedEevee notes, could expose “the AI’s developers [of the SummaryBot, which is used by the EA Forum Team] to liability for libel if the claims are false.” In fact, every single claim from (1) to (4) is false, and (5) doesn’t apply because I haven’t misrepresented anyone—my articles are full of citations, hyperlinks, quotes, and block quotes for a reason.3 So, the EA community has now made themselves liable for libel, it seems. Here is the offending comment:
The EA News Twitter account also retweeted the article, which might also increase the community’s liability for libel:
I have given the EA community every opportunity to rectify this situation. In good faith, I reached out to the Centre for Effective Altruism, though this was before I knew that they had violated confidentiality promises in the past. I also noticed that “X” deleted their EA Forum account entirely just before the Mark Fuentes article was posted. I wonder why. END OF EDIT.]
4. Appendix: Misrepresentations by Andreas Kirsch
On February 2, 2023, an Oxford DPhil student named Andreas Kirsch posted a Twitter thread in which he claims that the “Mark Fuentes” “article checks out.” He says that it “is not defamatory bc it’s substantially true,” and that “all the claims seem well evidenced: on the scholarship, but also on the other things.” He then tagged three employers of mine.
In truth, Kirsch did not verify the article. Much of the information in the previous subsection will be new to him, and several exchanges via DM after he published his thread indicated that he misremembered or conflated parts of a brief discussion we had prior to him posting the thread. When pushed, on Twitter, about what he did to verify the “Mark Fuentes” article, Kirsch declined to offer any evidence to support his claims, or to say who the supposed “sources” were that he contacted. Instead, he asked his social media followers, and the employers of mine who he tagged, to simply take his word for it.
In more detail:
(1) If Kirsch’s “sources” were people like Pluckrose, Lindsay, and Boghossian, this would be like asking the school bully if they ever hit anyone. People are more than welcome to believe a bunch of neo-fascist Internet trolls over me if they’d like. (In our exchanges, Kirsch admits to not knowing who these people are. But he should have researched them before claiming to have verified the “Mark Fuentes” article. Pluckrose and Lindsay, for example, have played an integral role in launching the moral panic over “critical race theory” in the US; Lindsay is a Trump-supporter who was permanently banned from Twitter for racism, transphobia, and popularizing the term “ok groomer,” part of a conspiracy theory that LGBTQ people like myself are “grooming” children. Boghossian has defended Nazis against being called Nazis, wondered why gay people should be “proud,” and moved to Hungary because he’s played a part in Victor Orban’s country-wide ban on gender studies in universities. In contrast, much of my public-facing writing since 2015 has focused on promoting social justice issues.)
(2) Kirsch writes: “I have also asked @xriskology [my Twitter handle] to comment & heard their side and their explanations.”
This is misleading. Kirsch asked me a handful of questions, I didn’t respond in much detail, he never followed-up, and he never asked me for references to confirm that, for example, Pluckrose has lied about me. I sent him a short email discussing the people above, which he apparently ignored. This is not verification.
(3) Kirsch writes the following:
This is false. He did not ask me for screenshots. The only reference he made to screenshots in our DMs was the following (quoting him): “A substack would be good to explain your perspective. Esp if you had a screenshot of that tweet from the next day, that would go a long way to refute the claims.” Kirsch has misremembered. This is something he should have double-checked before publishing his thread.
(4) Kirsch writes:
I did not say that I “deeply regret making a racist ‘peurile’ joke on Twitter years ago” because I never made a racist joke on Twitter, nor did I suggest to Kirsch that I “regretted” this. It became clear in a subsequent DM that Kirsch had conflated two separate things I told him. Here’s what I actually wrote regarding the completely false accusation from “Mark Fuentes” that I once made a racist remark in early 2018:
Kirsch, apparently, misinterpreted the above as me (a) admitting that I made a racist comment (I didn’t), and (b) stating that I regret having done so (doesn’t apply). Clearly that is not what I told him.
Kirsch has yet to publicly acknowledge that he misremembered and conflated messages I sent him, and messages that he’d sent to me, and he has not provided any evidence that he actually verified the defamatory “Mark Fuentes” article. (UPDATE: Kirsch has now started sharing screenshots of private messages I sent him while continuing to tag my employers.)
Note, however, that MIRI was not directly involved in this workshop.
Why was Shermer’s name deleted? Presumably because I accused him of being a rapist. Why would I do this? Because I befriended his rape victim and believe her.
For a response to one of Mark Fuentes’ claims about me, which is entirely false (and Mark Fuentes knows this), see here: https://docs.google.com/document/d/1Y-QAb9MT3tykldK0W-KBtJKVVbUnUkcSFxcjslCZq5A/edit?usp=sharing.