Key Moments
How to lose $35 Billion Dollars Betting on AI
ColdFusionWant to know something specific about what's covered?
We've already dissected every moment. Ask and we will deliver (with timestamps).
Key Moments
A 24-year-old hedge fund manager lost $35 billion betting on AI's rapid ascent, only to be wiped out by market volatility and excessive leverage.
Key Insights
Leopold Ashenbrenner's hedge fund, Situational Awareness LP, managed $45 billion but lost approximately $35 billion in two weeks due to concentrated AI stock investments and short positions that simultaneously declined.
Ashenbrenner's strategy relied heavily on AI infrastructure stocks and shorting software companies, a 'hedge' that effectively doubled his exposure to the same AI narrative, amplified by 400% leverage.
Despite the catastrophic loss, Ashenbrenner invested another $400 million into a private company days after the market crash, demonstrating extreme conviction despite devastating risk management failures.
The core of his strategy was a belief that AGI would arrive by 2027, a thesis that led him to focus on highly volatile public markets for rapid profit rather than venture capital, a decision that proved detrimental.
Institutional investors in New York, unlike their Silicon Valley counterparts, questioned Ashenbrenner's lack of a contingency plan if his AI thesis was wrong or delayed, highlighting a disconnect in risk perception.
The collapse of Situational Awareness LP led to a massive liquidation of assets, with Citadel acquiring $16 billion of the fund's stock portfolio at a discount in an overnight auction to cover collateral demands.
A spectacular wedding week and a $35 billion implosion
The story begins with a dramatic paradox: Leopold Ashenbrenner, a 24-year-old, was celebrating his wedding in Carmel, California, surrounded by Silicon Valley elites. Just two years prior, he had launched a hedge fund with $225 million, which had ballooned to an astonishing $45 billion under management. However, on his wedding week, his biggest investments began to collapse, leading major banks like Goldman Sachs, JP Morgan, and Bank of America to demand more collateral. Within days, approximately $35 billion of his empire vanished, a staggering sum equivalent to spending $1 million every day for nearly a century. This catastrophic loss occurred simultaneously with his wedding, making it an unforgettable, albeit disastrous, personal and professional milestone.
The prophecy of AGI and the Situational Awareness essay
Ashenbrenner's reputation was built on his perceived ability to predict the future, particularly in AI. In June 2024, he published a 165-page essay titled 'Situational Awareness: The Decade Ahead.' Its central thesis was that Artificial General Intelligence (AGI) would emerge by 2027, a timeline the world was unprepared for. This essay went viral within Silicon Valley, attracting significant investor attention. The essay itself was born from his time at OpenAI, where he worked on the super alignment team. Although he was later fired from OpenAI, his official reason being the sharing of a brainstorming document, Ashenbrenner claims his troubles began after he sent a memo to the board warning about inadequate security against foreign actors stealing company secrets. This memo was reportedly deemed 'unconstructive' by HR.
From FTX to OpenAI: a controversial trajectory
Leopold Ashenbrenner's early career was marked by connections to high-profile, and in some cases, problematic tech companies. He graduated from Columbia University at 19 and his first job was at Sam Bankman-Fried's FTX, a crypto exchange that later collapsed due to fraud. Following FTX, he moved to OpenAI, working on critical AI safety initiatives. His tenure there was short-lived; he was fired in April 2024. Ashenbrenner attributes his firing to internal disagreements over security concerns, a narrative that contrasts with OpenAI's official explanation. These early experiences, while brief, placed him at the heart of major tech industry events and controversies.
The investment pitch and the 'alpha' of AI prediction
Ashenbrenner's hedge fund, Situational Awareness LP, was pitched as a direct play on his AGI thesis. The core argument was that if AGI was coming rapidly, traditional market analysts were not pricing it in correctly. He believed a select group of individuals, particularly in San Francisco, possessed 'alpha' – a superior ability to foresee market-moving trends. He argued that the market's underestimation of compute and energy demands for AI development meant that analysts focused on conventional metrics should, by economic principles, be bankrupt. This perspective led him to question why governments hadn't nationalized AI labs and why society hadn't fully integrated the implications of advanced AI, framing his fund as a way to capitalize on this market mispricing.
Silicon Valley wunderkind meets Wall Street skepticism
The fund attracted significant backing from prominent figures like the Collison brothers (Stripe), former GitHub CEO Natt Friedman, and investor Daniel Gross. Ashenbrenner was described as intense, scarily smart, and brash, with an aura of a 'wonderkind.' However, his approach also drew criticism; some suggested his ideas were repackaged existing AI industry thoughts. While Silicon Valley was captivated, Wall Street investors were more cautious. When seeking institutional capital in New York, a wealthy investor reportedly asked Ashenbrenner what would happen if his AI thesis was wrong or delayed, a question to which Ashenbrenner apparently had no detailed answer, relying instead on his conviction. This highlights a divergence in risk tolerance and foresight between the two financial hubs.
The strategy: AI infrastructure longs and software shorts
Ashenbrenner's strategy was twofold: buying stocks of companies building the physical backbone of the AI boom – chipmakers, power companies, and cloud providers (e.g., SK Hynix, Coreweave, Nebulas) – and simultaneously shorting software companies (e.g., Adobe, Salesforce) that he believed AI would make obsolete. This 'long-short' approach was intended as a hedge. However, both sides of his trades were underpinned by the same core belief: that AI would rapidly upend the entire economy. Crucially, he employed 400% leverage, meaning for every dollar of his own capital, he borrowed four dollars. While this strategy yielded spectacular returns for about a year, it was inherently vulnerable to synchronized downturns in both his long and short positions, especially under extreme leverage.
Market 'ructions' and the $35 billion margin call
The downfall began in July 2026 when whispers of Chinese open-weight AI models performing comparably to leading Western models at a fraction of the cost spooked the American markets. This news triggered a sharp decline in Ashenbrenner's concentrated long positions, with the six biggest holdings falling between 50% and 78% from their peaks. Simultaneously, the software stocks he had shorted unexpectedly rose. This meant both halves of his supposed hedge were losing money, a scenario that a true hedge is designed to prevent. His prime brokers demanded massive collateral increases, which Ashenbrenner couldn't meet. This triggered a forced liquidation of his assets. Approximately $16 billion of his stock portfolio was sold overnight at a significant discount to Citadel, one of the most successful hedge funds, which won the auction. The fund's value plummeted from $45 billion to $10 billion, resulting in the $35 billion loss.
Beyond the balance sheet: conviction and continued investment
Despite losing $35 billion and facing immense pressure from lenders, Ashenbrenner's fund survived due to an estimated $5 billion stake in the privately held, illiquid company Anthropic, where his future wife was Chief of Staff to the CEO. What's remarkable is that days after the Citadel deal, Ashenbrenner wired $400 million into an unnamed private company he had previously invested $100 million in, showing an astonishing level of conviction. This highlights a critical distinction in finance: conviction and risk management are separate disciplines. His extreme leverage, combined with a strategy that effectively doubled down on a single AI narrative, created a fragile position. As noted by industry veterans, investors often use a fraction of the 'Kelly criterion' (a formula for optimal bet size) to preserve capital, sacrificing theoretical maximum returns for survival. Ashenbrenner's experience serves as a stark reminder that even a correct long-term thesis can be invalidated by poor timing and excessive leverage.
Mentioned in This Episode
●Products
●Software & Apps
●Companies
●Organizations
●Books
●People Referenced
Leopold Ashenbrenner's Fund Performance and Collapse
Data extracted from this episode
| Metric | Value | Timestamp/Period |
|---|---|---|
| Initial Capital | $225 million | Less than 2 years before collapse |
| Peak Assets Under Management | $45 billion | Before collapse |
| Loss Amount | $35 billion | In 2 weeks |
| Final Assets Under Management | $10 billion | After collapse |
| Leverage Ratio | 4:1 | At time of collapse |
| Largest Long Positions Drop | 50-78% | End of July 2026 |
| Stake in Anthropic | $5 billion | During collapse |
| Investment in unnamed private company | $500 million | Days after Citadel deal |
Common Questions
Leopold Ashenbrenner is a 24-year-old hedge fund manager who founded Situational Awareness LP. His fund collapsed, losing $35 billion, due to an aggressive investment strategy heavily reliant on leverage and a concentrated bet on AI stocks, which backfired when the market turned.
Topics
Mentioned in this video
A 24-year-old hedge fund manager who lost $35 billion betting on AI. He founded the hedge fund 'Situational Awareness LP' and previously worked at FTX and OpenAI.
Founder of FTX, where Leopold Ashenbrenner had his first job outside of college.
Host of The Tim Ferriss Show, who featured Leopold Ashenbrenner and called him the 'Nostradamus of AI'.
Investors in Leopold Ashenbrenner's fund, known for founding Stripe.
An investor in Leopold Ashenbrenner's fund.
Economist who called Leopold Ashenbrenner an 'economics prodigy' at age 17.
Quoted in the video discussing the concept of accelerating intelligence in machines, a concept relevant to AGI discussions from 1996.
A wealthy New York investor who reportedly questioned Leopold Ashenbrenner about what would happen if his AI thesis was wrong.
Soon-to-be wife of Leopold Ashenbrenner, serving as Chief of Staff to the CEO at Anthropic.
Hedge fund manager who commented on the Leopold Ashenbrenner saga, referencing the 'Kelly rule' of investing.
The hedge fund founded by Leopold Ashenbrenner, named after his essay predicting AGI. It managed $45 billion before collapsing and losing $35 billion.
University where Leopold Ashenbrenner graduated as valedictorian at age 19 in 2021.
A fraudulent multi-billion dollar crypto exchange founded by Sam Bankman-Fried, where Leopold Ashenbrenner worked before its collapse in late 2022.
A successful hedge fund that bought $16 billion of Leopold Ashenbrenner's stock portfolio in an auction after his fund's collapse.
One of the prime brokers for Leopold Ashenbrenner's fund, which demanded more money to cover losses, contributing to the fund's collapse.
One of the prime brokers for Leopold Ashenbrenner's fund, which demanded more money to cover losses, contributing to the fund's collapse.
AI research company where Leopold Ashenbrenner worked on the super alignment team before being fired in April 2024.
The company founded by the Collison brothers, who were among the investors in Leopold Ashenbrenner's Situational Awareness LP fund.
A firm that reportedly met with Leopold Ashenbrenner but passed on investing, showing initial Wall Street skepticism.
A company whose stocks were part of Leopold Ashenbrenner's long positions. It experienced a significant price drop.
A software company that Leopold Ashenbrenner shorted, believing AI would make its enterprise tools redundant.
A software company that Leopold Ashenbrenner shorted, believing AI would make its enterprise tools redundant.
An AI company where Leopold Ashenbrenner had a $5 billion stake, which helped his fund survive the collapse as it was privately held and illiquid.
A stock holding of Leopold Ashenbrenner's fund that plunged 57% in under a month, significantly contributing to losses.
A 165-page PDF essay published by Leopold Ashenbrenner in June 2024, arguing that AGI was coming by 2027 and the world was unprepared. It went viral in Silicon Valley.
Publication that ran a deep profile on Leopold Ashenbrenner in October 2025, painting a picture of a divisive figure.
More from ColdFusion
View all 90 summaries
23 minWhy Kids Can't Read Anymore
30 minWhat Does Palantir Actually Want? (Re-upload)
28 minWhy Building AI Data Centres Isn’t Working Anymore
30 minWhat Does Palantir Actually Want?
Ask anything from this episode.
Save it, chat with it, and connect it to Claude or ChatGPT. Get cited answers from the actual content — and build your own knowledge base of every podcast and video you care about.
Get Started Free