Leopold Aschenbrenner built a $45 billion fund called Situational Awareness. Named it after his own essay. Lost most of it within days.
The man left OpenAI to manage money based on predicting AI timelines. Turns out predicting when AGI arrives does not help you predict what happens Tuesday. Different skill sets.
Forty-five billion dollars. Gone in days. Not months. Not a slow bleed where you can pretend it's strategic repositioning. Days. The kind of loss that makes you wonder if anyone was actually watching the screens or if they were all in a Slack channel debating whether GPT-7 would be conscious.
Aschenbrenner wrote extensively about situational awareness in AI systems. The ability to understand context and adapt to changing conditions. His fund apparently lacked both. Named your hedge fund after the one thing you failed to demonstrate. Bold.
Retail traders saw "AI hedge fund" and probably thought it meant the fund itself was intelligent. It was not. It was a regular fund that lost money at AI speed. Forty-five billion to almost nothing faster than you can say "but the technical indicators."
The essay was about AI understanding its position in the world. The fund's position was apparently "about to collapse" and nobody mentioned it in the morning meeting. Situational awareness would have been useful there.
Former OpenAI researcher applies cutting-edge AI insights to financial markets. Loses everything immediately. Markets remain unimpressed by your ability to write long PDFs about superintelligence. They want boring things like risk management and not lighting billions of dollars on fire.
Somewhere a day trader with $3,000 in Robinhood is reading this and thinking he could have done better. He is wrong but he will never know it.
Photo by Marwen Larafa on Unsplash
Leave a Comment