Leopold von Aschenbrenner: The New Sam Bankman-Fried?
Plus, Dario Amodei’s wife tried to get Jeffrey Epstein to fund her "luxury porn" company. (2,800 words)
Cover photo is from here.
Summary1:
A woman named Cami Clark, who hung out with EA-longtermists and married one of the most powerful EA-longtermists in the world (Dario Amodei), exchanged emails with Jeffrey Epstein after his 2008 conviction in hopes of getting him to fund her porn company, whose tagline was: “intellectually promiscuous.”
Another leading EA-longtermist, Leopold Aschenbrenner, who worked for Bankman-Fried’s FTX Future Fund and married a former colleague at the Future Fund, Avital Balwit — who now works for Clark’s husband, Amodei — started a hedge fund that invested heavily in Anthropic and recently lost $35 billion of its $45 billion AUM.
Lol.
Dario Amodei’s Wife Tried to Get Jeffrey Epstein to Fund Her “Luxury Porn” Company
Last February, I wrote a Truthdig article about Jeffrey Epstein’s extensive connections to the transhumanist community. Epstein himself was a transhumanist and eugenicist (but I repeat myself) who wanted to have his head and penis cryogenically frozen and aimed to create a “super-race of humans with his DNA” by impregnating up to 20 women at a time on his Zorro Ranch.
He was in contact with Bryan Johnson, and appears to have met Aubry de Grey (the longevity researcher) over Skype. He was also pals with Peter Thiel and Peter Diamandis, and exchanged emails with Elon Musk, who practically begged Epstein to let him visit the pedophile and sex trafficker’s home-base in the Virgin Islands.
Epstein was also long-time buddies with Ben Goertzel, who congratulated Epstein when he emerged from prison in 2009, convinced Epstein to transfer $50,000 to Eliezer Yudkowsky’s Singularity Institute (now the Machine Intelligence Research Institute), and wrote the following to Epstein after sending him birthday wishes:
Now if you want to have 100000 more birthdays you should fund my AI project so we can create a thinking machine that will solve the problem of mortality!!!
These are just a few examples of Epstein’s connections to the TESCREAL movement.
***
It turns out that Epstein was also friendly with Cami Clark, the wife of Anthropic CEO Dario Amodei. Previously, Clark had dated Eric Schmidt, the former Google CEO who claimed we should build ASI (artificial superintelligence) despite the environmental costs. She married Amodei in 2022 (having been married once before to an architect who was 64 years old while she was just 20).
As Forbes reports:
Her previous attempted businesses include founding a self-described “free luxury porn” company that catered to women and a “social dieting app.”
Clark met Epstein in March 2011 through John Brockman, a literary agent who knew Epstein, with emails in the Epstein files showing Brockman telling Epstein he “should connect for dinner with my girls,” referring to Clark and her business partner Michelle Capocefalo. [The full quote is: “You should connect for dinner with my girls — Cami and Michelle. They’re in LA raising money for porn movie aimed at women’s market.”]
“We would love to have dinner with you this evening,” Cami said in an email responding to Brockman’s introduction, then sending an email the next day that passed along her company’s funding requirements and exclaimed, “It was FANTASTIC meeting you last night!”
Clark then asked Epstein in a March 2012 email exchange about whether he’d be interested in investing in her and a colleague’s businesses, though there’s no indication in the Epstein files that the financier made any investment.
The porn company was apparently called Eddice. It “described itself as a ‘revolutionary porn company’” that emphasized “sex positivity in a male-oriented porn industry, and carried the tagline: ‘intellectually promiscuous.’”
All of this happened after Epstein pled guilty to soliciting prostitution from a minor and spent 13 months in prison — out of the 18 months he was sentenced to in a “sweetheart deal.” Upon being released, he almost immediately began raping an 18-year-old woman who had “recently arrived from Uzbekistan.”
As with Noam Chomsky, who became close friends with Epstein, Clark appears not to have cared that Epstein was a registered sex offender.
Forbes continues:
After their initial introduction through Brockman, Clark sent Epstein the script and treatment for her company’s first films a few days later in March 2011, noting she thought Epstein “and the ladies” might enjoy them and saying the content was “a little nsfw.” Clark’s March 2012 exchange with Epstein is the lengthiest conversation between them that appears in the Epstein files, with Epstein initially asking, “i’m sorry, who is this,” after Clark reached out and asked to put Epstein in touch with another female entrepreneur launching a lingerie line. When Clark reminded him that she has the “free luxury porn company” and asked if that “ring[s] a bell,” Epstein responded, “yes, a loud gong.” (“I was going to say, you would be the first person ever that didn’t remember us :),” Clark responded.) … Epstein then asked what happened to Clark’s business plans with the porn company and she said they were moving “full speed ahead,” pitching Epstein on investing in her company’s foray into video. “Can’t do TV sex,” Epstein responded, to which Clark said, “I didn’t think so :).”
Forbes adds that “other emails in the Epstein files show Clark keeping the financier on her email list — inviting him to a housewarming party in July 2012 — and that Epstein connected with her on LinkedIn a year later in 2013.”
***
Clark has almost no online footprint. She keeps a low profile, but appears to be influential behind the scenes, even playing a role in getting Anthropic early funding. As the Wall Street Journal, which broke the story, observes:
When Indian Prime Minister Narendra Modi invited AI leaders to a meeting in New Delhi earlier this year, security protocols allowed each executive to bring one additional person with them. Most brought colleagues, but Anthropic CEO Dario Amodei brought his wife, Cami Clark.
Clark doesn’t work at Anthropic, but she is often seen sitting in the front row while Amodei talks at events such as Davos or can be found chatting up investors at gatherings such as the Allen & Co. conference in Sun Valley.
She acts as a sounding board and strategic adviser for Amodei, according to people close to the company. She also brought Anthropic a key early investor, former Google CEO Eric Schmidt — whom she had dated — as it was getting off the ground in early 2021, some of the people said.
WSJ also notes that
Clark spent time at the group house in San Francisco where Amodei, his friend Holden Karnofsky and his sister, Daniela — who would later marry Karnofsky — lived with other key figures from the AI safety and Effective Altruism worlds. The movement, centered largely in Silicon Valley, aims to use data and reason to determine how to do the most good possible.
And:
Clark has recently told people involved in politics that the company sees its mission as protecting America and isn’t as “woke” as its detractors might think it is, according to people familiar with the matter.
At the Sun Valley conference in July, where she spent much of the week by her husband’s side, she had lunch with Ivanka Trump, a friend, and chatted with Jared Kushner, whom Amodei had approached earlier in the year to invest, people familiar with the matter said.
She strikes me as a hyper-ambitious person with low moral standards, which of course describes many power-hungry vultures in Silicon Valley — including her husband. I wonder who else in the TESCREAL/ASI race has had connections with Epstein that haven’t yet been reported.
Womp Womp: Situational Awareness Loses $35 Billion
Leopold von Aschenbrenner — a German-born investor and AI safety researcher whose name I jokingly add the aristocratic “von” to — is an Effective Altruist (EA) and longtermist who gained a bit of Silicon Valley fame for his wildly hyperbolic 2024 essay, “Situational Awareness.” In it, he hyperventilates about ASI being imminent and argues that it’s very important for the US to win the ASI race against China.
Here’s what I write about his essay in my forthcoming book on the TESCREAL ideologies:
[Aschenbrenner] contends that “China is still competitive in the AGI race, and China being first to superintelligence would be very bad because it may enable a stable totalitarian world regime. So the US must win to preserve a liberal world order.” The obvious problem is that the US isn’t much better than China right now. The current US president is a twice-impeached convicted felon who embraces fascist authoritarianism. He inspired a mob of MAGA supporters to perpetrate an insurrection in hopes of preventing the peaceful transfer of power in 2024 — essentially, the goal was to overthrow the US government.
Xi Jinping in China would no doubt strive to implement a totalitarian state. Aschenbrenner is right about that. But so would Donald Trump. China engages in mass surveillance of its citizens, but so does the US. The AI companies themselves are complicit in this, as Anthropic is working with Palantir, a mass surveillance company, while OpenAI is okay with the Pentagon using its AI models for the purposes of mass domestic surveillance.
Hence, the geopolitical race to AGI isn’t actually between democracy and authoritarianism. It’s between two variants of authoritarianism: US-style fascism and Chinese-style communism. Whichever country builds ASI first, the outcome is bound to be bad for everyone. Ironically, Aschenbrenner appears to be situationally unaware of this fact.
Aschenbrenner’s “impressive” CV includes a stint at Sam Bankman-Fried’s cryptocurrency exchange platform, FTX. Specifically, he worked for the FTX Future Fund, run by Nick Beckstead — the longtermist who argued that we should prioritize saving the lives of people in rich countries. The Future Fund also gave $500 million in funding to Anthropic.
While still a teenager, he wrote a paper titled “Existential Risk and Growth,” and cofounded the Effective Altruism club at Columbia University when he was a student there. He spent time at the Future of Humanity Institute (before it was shut down due in part to a racist email from it’s founder that I uncovered in late 2022), as well as Will MacAskill’s Global Priorities Institute. He later joined OpenAI’s “Superalignment” team, but left after OpenAI accused him of sharing confidential documents (which von Aschenbrenner disputes).
Following his viral-in-the-Valley 2024 essay, he founded a hedge fund of the same name: Situational Awareness, which he describes as “an investment firm focused on AGI.” At its peak, Situational Awareness had $45 billion assets under management (AUM), but last month it lost a catastrophically huge $35 billion. That still leaves an eye-watering $10 billion behind, but the massive drop in assets reminded many of the ignominious collapse of Aschenbrenner’s former employer: Bankman-Fried.
As CNBC reports,
critics pointed out that Aschenbrenner had no experience running money prior to launching his fund in July 2024, calling him more lucky than smart. Some noted that his early work experience was at the doomed crypto firm FTX, where he helped now-disgraced founder Sam Bankman-Fried run a charity out of a Bahamas penthouse.
“A lot of people saw this blow-up as a matter of not if, but when,” said Jerry Diao, who runs a Wall Street coaching firm. “Maybe his views on AI are correct in the long run, but in the public markets, you have to be prepared for the short-term.”
The same article opens with the rather sardonic line: “Two years ago, Leopold Aschenbrenner argued he was one of few people in the world who saw the future clearly.” Now, von Aschenbrenner has made himself the butt of jokes and thrust EA into the spotlight once again. (Aschenbrenner was essentially earning to give, the very same career path that led Bankman-Fried into crypto.)
As my friend David Z. Morris writes in a quite amusing article, “Son of Sam: Aschenbrenner was SBF 2.0”:
An infamous Effective Altruist thought experiment is “the coin flip to end the world.” Sam Bankman-Fried interpreted this as an injunction to take catastrophic risk if the average “expected value” of winning a long bet was even slightly positive.
That high-risk mindset also seems to have ruled at Situational Awareness, a huge, AI-focused hedge fund that exploded in blood and guts last week. It was run by a deeply involved Effective Altruist named Leo Aschenbrenner.
Aschenbrenner’s downfall very closely tracks that of Sam Bankman-Fried. Specifically in that they were both young, loved using very high leverage in volatile markets, believed trend-lines only go up, and apparently thought risk management is for cowards.
Morris continues:
Aschenbrenner still got caught by the exact same ratchet that killed SBF.
Situational Awareness LP grew incredibly fast after 2024, to a peak of $45 billion in AUM, by using high leverage to bet on the continued rapid growth of AI. Aschenbrenner had kept making the same bet for years, at larger and larger size, to the point where his leverage may have been propelling the broader market.
Then July was choppy for AI stocks — incredibly, even $Nvidia was flat - and that was all it took to destroy Situational Awareness. …
[T]he linear, predictionist modelling that rules EA, alongside the idea that its adherents are uniquely genetically gifted, lead believers to take on extreme risk and, inevitably, to blowups like Aschenbrenner’s.
He further notes that, almost comically,
Aschenbrenner’s unluckiest investors lost something like 110% of the current mark-to-market value of their assets.
That means Leo Aschenbrenner was far, far worse for many investors than Sam Bankman-Fried was for FTX depositors.
***
Of note is that von Aschenbrenner’s wedding to Avital Balwit happened just weeks after the $35 billion loss. Balwit is Dario Amodei’s chief of staff at Anthropic and a former employe at Bankman-Fried’s FTX Future Fund — just like her husband was. Note also that Situational Awareness had a multi-billion-dollar investment in Anthropic. As Morris observes, “the largest remaining asset in Situational Awareness LP portfolio is reportedly a large stake of illiquid Anthropic equity.” (This community is so incestuous!)
Balwit, I should add, wrote a long response back in 2021 to a couple of articles I published critiquing longtermism — an ideology that both she and her husband obviously subscribe to. Her critique of my critique is pretty abysmal, spurring me to write a response in which I dissect each paragraph, showing how far her criticisms miss the mark. I had some personal communication with her because I sent her my response prior to publication for comments. (A courtesy I always extended to EAs and longtermists, though they never reciprocated.)
In a parallel universe, where I wasn’t moved by considerations of moral and intellectual integrity to abscond from the TESCREAL movement, it’s entirely possible that I, too, would be working at Anthropic, making perhaps half-a-million dollars or more per year. It’s pretty wild to think. As one prominent EA noted on social media: “The only kinds of groups I know of more predictive of wealth than ‘reading LessWrong in 2015’ is like ‘holding bitcoin in 2011’ or ‘working at a hedge fund.’” That was me: reading LessWrong in 2015!
In other words, the TESCREAL community takes care of its own. There are tens of billions in funding, and money gets around. Although my departure from the movement was inevitable — given that I do care about integrity — it’s wild to think about where I might be if I’d remained, especially given that I was once listed as the 6th most prolific contributor to the academic literature on existential risks!
This is precisely why — if it’s not too crass for me to mention here — you should consider supporting me on Substack, GoFundMe, or Patreon. There just isn’t much money going around for critics of the TESCREAL cult.
Anyways, the whole von Aschenbrenner debacle is rather hilarious. As I wrote years ago, EA is the gift that keeps giving to critics.
As always:
Thanks for reading and I’ll see you on the other side!
PS. In case you need a chuckle, I saw this on Facebook the other day. :-)
CORRECTION: In the original post, I accidentally wrote “$45 billion” on a few occasions instead of “$35 billion,” which is how much Situational Awareness lost. My apology for the error!
Not AI! :-)
In the words of Aries Spears: “Nuke the whole TESCREAL community”.