Situational Awareness Fund Crashed After 1,000% Run

Leo Aschenbrenner's Situational Awareness hedge fund, up more than 1,000% in two years, crashed in July and was forced to sell its public stocks at fire-sale prices to Ken Griffin after a fairly minor market correction exposed bets that were both highly correlated and highly leveraged.

As Theo Baker reported in The Atlantic, Aschenbrenner built a reputation as the "Nostradamus of AI" on the strength of his viral essay Situational Awareness. "I can see how AGI will be built," he wrote. Born in Berlin to two doctors, he graduated from Columbia as valedictorian at 19 and did stints at FTX and OpenAI before launching the fund. He seemed cocky and eccentric, but, as the source notes, "you couldn't argue with his results" — until the crash.

Why does confidence keep preceding the bust?

The essay draws on Martin Heidegger's The Question Concerning Technology, which describes technology as "a way of revealing": a hydroelectric dam doesn't create a river's energy, it channels it. When a new technology reveals new possibilities, confusion follows, and those who most confidently point the way forward attract investors, press quotes and conference invitations.

Earlier booms followed the same script. Early web companies Webvan, eToys and MP3.com reached huge valuations before crashing. Enron succeeded as an energy trader before trying to repeat that success in water and bandwidth markets. The problem wasn't that the ideas were bad — later versions of the same ideas succeeded — but that, in the source's framing, all ideas are wrong in some way, and a leader's job is to make the idea less wrong over time.

Research from the Federal Reserve Bank of St. Louis shows why the pattern repeats with AI. Productivity is soaring for some tasks, like those related to computers and math, but AI has far less impact on most work. People closest to the tasks enjoying a productivity boom easily mistake their experience for a universal one — and gain a certainty the rest of us don't share.

What is the halo effect costing investors?

Psychologists call it the halo effect: an impression in one area colors our judgment in another. In his book The Halo Effect, Phil Rosenzweig points to Cisco during the dotcom boom. When Cisco was flying high, its culture was described as unparalleled, with happy people working long hours and loving it. When the market tanked, that same culture came to be seen as "cocksure" and "naive."

A more recent case was DOGE, the Department of Government Efficiency spearheaded by Elon Musk, the entrepreneur behind PayPal, Tesla and SpaceX. Many cheered the idea that he could control the federal budget. The GAO found most of DOGE's savings claims were incorrect or unsupported. A study in The Lancet projected that continued USAID cuts could cause more than 14 million additional deaths by 2030. The US federal deficit and debt have surged.

The halo effect looms larger in tech than elsewhere, the essay argues: Nobel laureates like Jennifer Doudna (CRISPR) and Katalin Karikó (mRNA), whose work has arguably had more impact than anything out of Silicon Valley, aren't trying to run the federal government or weigh in on foreign policy.

How did the purity spiral hit Aschenbrenner?

Aschenbrenner benefited from the halo effect: admirers assumed his academic brilliance would transfer to hedge fund investing. Whatever the merit of his AI ideas, he failed to grasp basic principles of risk management.

Social psychologist Jonathan Haidt argues in The Righteous Mind that opinions often reflect team membership. Aschenbrenner wasn't just analyzing technology; he was signaling a techno-optimist identity. That identity signaling drives group polarization and "moral outbidding," producing a purity spiral where the most extreme views are proudly displayed and moderation becomes weakness. The result is what George Soros calls reflexivity: skepticism gets dismissed, leverage increases, prices stray from reality — until the gap becomes unsustainable and the boom becomes a bust.

What should Silicon Valley do now?

The essay's answer: build a red team. Before Operation Neptune Spear, the raid that killed Osama bin Laden in Abbottabad, Pakistan, President Obama ordered a separate group to review the evidence, check assumptions and look for alternative explanations rather than build more consensus.

Every boom-bust cycle has a breakthrough insight at its core — railroads and the internet really did transform economies, and AI promises to be no less consequential. But as marginal players pile in and push values past rational basis, reality eventually sets in. The Situational Awareness crash was the warning: in Silicon Valley, confidence itself is becoming the problem.