How to lose $35 Billion Dollars Betting on AI
ColdFusion · 2026-09-16 · 22 min
https://www.youtube.com/watch?v=7K_sA6o1dOEVideo summary
Leopold Aschenbrenner’s AI fund fell from $45 billion to $10 billion after 400% leverage magnified a two-sided bet gone wrong.
At 23, Leopold Aschenbrenner launched Situational Awareness LP after his essay predicted AGI by 2027. Backed by prominent Silicon Valley investors, the eight-person fund grew from $225 million to a reported $45 billion, buying AI infrastructure stocks and shorting software companies. But both sides relied on the same AI thesis, amplified by roughly four-times leverage. In July 2026, AI-linked holdings plunged while the shorted software stocks rose; lenders including Goldman Sachs, JPMorgan Chase, and Bank of America demanded collateral. Aschenbrenner’s fund sold about $16 billion in stocks to Citadel at a discount, shrinking to roughly $10 billion. A private Anthropic stake helped it survive. Days later, he invested another $400 million in an unnamed company. The story’s lesson: conviction is not risk management, and leverage can force investors out before a thesis has time to play out.
Chapters
- 0:00The $35 billion crisis: Leopold Aschenbrenner’s $45 billion fund collapsed during his wedding week
- 2:08From FTX to OpenAI: Leopold’s AGI-by-2027 essay and his FTX career
- 3:36Situational Awareness: OpenAI security warnings, the “Nostradamus of AI,” and a new fund
- 5:37The investment pitch: scaling curves, rising compute demand, and AGI mispriced by markets
- 6:48A divisive wunderkind: Columbia at 15, an eight-person fund, and investors questioning his AI thesis
- 9:31Sponsor: Wispr Flow
- 11:17The trading strategy: SK Hynix and other AI longs, software shorts, and 400% leverage
- 12:53Market collapse and exit: Chinese AI competition, a $16 billion Citadel sale, and the Anthropic stake
- 15:23The leverage trap: a $400 million private investment after 4:1 exposure and SanDisk’s 57% drop
- 17:22Risks of conviction: a 50% drawdown could liquidate Leopold before AGI arrived
- 19:14Lessons and fallout: Cliff Asness’s Kelly rule, JPMorgan’s cutoff, and renewed AI bets
This is a Tier 1 public summary
Whether the chapter key points, section summaries and mind map are public is up to the person who shared it. Want the full analysis?Submit one yourself.
More from this channel
The RAM Crisis Keeps Getting WorseColdFusionAI data centers are tightening RAM supplies, driving up device costs as Samsung, SK Hynix, and Micron prioritize enterprise buyers.
Yikes.ColdFusionColdFusion traces Claude’s role in military targeting, Anthropic’s refusal of surveillance terms, and OpenAI’s controversial Pentagon deal.
How The Internet’s Favourite AI Employee Went RogueColdFusionOpenClaw promised a capable AI assistant but exposed private data, compromised 4,000 developer machines, and helped drive a $1 billion mortgage fraud investigation.
OpenAI Kills Sora then Descends into ChaosColdFusionOpenAI shut down Sora after costs reached $15 million a day, while raising $122 billion and reportedly buying TBPN.
So... What Was the Point of Artemis II?ColdFusionArtemis II set a human distance record, survived a disputed Orion heat shield, and advanced NASA’s plans for lunar landings.
Related analyses
【阿格力】「最強多人操盤」的主動式ETF來了，00409A接棒00988A與00990A，全球型ETF在下半年繼續霸榜？阿格力博識台股回檔之際，影片看好台美日韓 AI 分工，並解析 00988A、00990A 與三經理人 00409A。
Claude 最強模型 Fable 5 深入解析：打著安全旗號，其實在搞反競爭？ | S2E61矽谷輕鬆談 Just Kidding TechClaude Fable 5 長任務明顯進步，卻頻繁誤降級 Opus 4.8，Anthropic 的安全與透明度引發質疑。
The Case for AI That Improves Itself | Deep Dives with a16za16z Deep DivesMirendil aims to cut AI research teams from 200 people to one and accelerate science, including Alzheimer’s research.
AI巨头们之间的资本混战，到底是个什么情况？小Lin说微軟持有 OpenAI 約 27% 股權，亞馬遜與 Google 轉向雙押 Anthropic，AI 投資也綁定天價算力訂單。
Big Tech is Acting Dodgy At the MomentColdFusionColdFusion warns AI spending is distorting Big Tech profits, with Google’s $200 billion Anthropic deal and Nvidia’s disputed $5.4 billion GPU sale raising concerns.