According to a report by CNBC yesterday, the AI hedge fund “Situational Awareness,” founded by former OpenAI researcher Leopold Aschenbrenner, is facing a severe crisis, with its assets shrinking from $45 billion (approximately 304.42 billion yuan at the current exchange rate) to about $10 billion (approximately 67.649 billion yuan at the current exchange rate).
Two years ago, Leopold said that he was one of the few people in the world who truly understood the future. He published a 165-page paper that briefly became required reading in Silicon Valley, and he himself came to be regarded as a prophet of the “superintelligence era.”
However, his foresight has faced a serious test since the start of this week. The AI hedge fund “Situational Awareness” that he operates has encountered a severe crisis amid a plunge in semiconductor stocks and pressure from margin calls on Wall Street. Earlier this month, the fund’s assets had reached $45 billion.
According to people familiar with the matter, Leopold was forced to sell all of his leveraged positions in stocks such as SK Hynix and CoreWeave to Ken Griffin of Citadel, shrinking the fund’s holdings to $10 billion.
Leopold’s rapid rise and dramatic downfall have drawn considerable attention from Wall Street and the technology sector, making him one of the most prominent failure cases of the AI boom.
Some critics pointed out that Leopold had no experience managing money before establishing the fund in July 2024. Early in his career, he joined the now-defunct cryptocurrency exchange FTX and helped the now-infamous Sam Bankman-Fried operate philanthropic projects at his penthouse mansion in the Bahamas.
Meanwhile, some traders and investment banking professionals pointed out that “Situational Awareness” used leverage as high as 400%, making its collapse unsurprising. Jerry Diao, head of a Wall Street investment training company, said: “Many people believe that this blowup was only a matter of time. Perhaps his judgment about the long-term trend of AI was correct, but when investing in public markets, you have to be prepared for short-term volatility.”
In a particularly dramatic twist, Leopold is engaged to Avital Balwit, chief of staff to Dario, the CEO of Anthropic, and plans to hold their wedding this weekend.
Leopold was born in Germany, and both of his parents are doctors. He displayed exceptional talent in mathematics and computer science from an early age, skipped grades several times, and graduated from high school at 15. After entering Columbia University, he attracted attention for an academic paper titled Existential Risk and Growth.
According to his classmates, Leopold was somewhat childlike and appeared socially awkward.
After graduating in 2021, he joined Future Fund, a charitable organization under FTX. In 2023, he joined OpenAI’s Superalignment team, where he researched how to align AI systems with human interests under the leadership of Ilya Sutskever.
After a hacker broke into OpenAI’s internal systems, he submitted a memo to the board warning that the company’s security measures were insufficient to withstand foreign espionage. In 2024, OpenAI fired him, citing the improper sharing of confidential information.
A few weeks after leaving OpenAI, Leopold published a paper arguing that artificial general intelligence (AGI) could arrive within a few years and that governments around the world had seriously underestimated the pace of AI development. In July 2024, he began raising seed capital for his own hedge fund. It had previously been rumored that he had secured a $250 million investment (approximately 1.691 billion yuan at the current exchange rate).