Situational Awareness LP

Leopold Aschenbrenner

Leopold Aschenbrenner runs Situational Awareness LP, the San Francisco hedge fund he started in 2024 around a single conviction: that artificial general intelligence arrives this decade, and the biggest winnings sit not in the model labs but in the physical buildout that powers them. A German-born math prodigy who skipped grades, graduated high school at fifteen, and finished Columbia as valedictorian at nineteen, he joined OpenAI's Superalignment team after school. In April 2024 the company fired him, saying he leaked confidential information — a claim Aschenbrenner disputes, insisting his dismissal followed an internal memo warning that OpenAI's infrastructure could not protect its model weights from foreign adversaries. Two months later he published "Situational Awareness: The Decade Ahead," a widely read essay arguing that compute scaling would carry AI past human capability far sooner than most believed. The fund takes its name from that essay. The fund is the essay turned into a portfolio. Instead of a diversified book, it runs a concentrated, high-conviction set of bets on the companies that must exist before AGI can: chipmakers, memory and storage firms, cloud and data-center operators, and the power generators that will run them all. Aschenbrenner filed his first 13F for Q4 2024 and quickly became one of the most closely watched names in the AI-infrastructure trade, wagering that the suppliers of compute — not the companies building the models — were the mispriced winners of the scaling boom.