When tech investor Jason Calacanis recently suggested on a podcast that Russia's invasion of Ukraine could be settled with a few land swaps and that the U.S. government has no business regulating mergers because it cannot predict the future, he was confidently offering simple answers to extraordinarily complex problems. Nobody with deep expertise in Eastern European history or antitrust law would make such claims so casually. Yet there he was, embodying a familiar pattern: highly successful people who assume their intelligence in one domain automatically transfers to every other.
The dynamic is not new. Bethany McLean's classic book, The Smartest Guys in the Room, chronicled the spectacular rise and collapse of Enron, a company that genuinely earned its reputation for brilliance early on. As someone who once worked on a natural gas trading desk, I saw firsthand how top-notch the Enron operation was. Their fatal flaw was not a lack of talent but arrogance. That hubris led them to place enormous bets on businesses they barely understood, from water and broadband to advertising and shipping, convinced that their energy-trading success had uncovered a universal business model.
That same overconfidence is visible in Silicon Valley today. In her book Careless People, former Meta executive Sarah Wynn-Williams describes executives so wealthy and powerful that they have lost touch with ordinary realities. She recounts a senior leader's genuine surprise upon learning that people living in refugee camps do not have jobs, during a discussion about how much to charge refugees for internet service. The anecdote captures how insulated success can distort basic understanding of the world.
Calacanis's comments on Ukraine fit the same mold. Had he wanted to grasp how Vladimir Putin himself justified the invasion, he could have read the 6,000-word essay the Russian leader published months beforehand. In that text, NATO is mentioned only twice and only in passing, while medieval Rus, Bohdan Khmelnytsky, and the Polish-Lithuanian Commonwealth appear roughly half a dozen times each. Understanding the significance of those references requires some knowledge of Eastern European history, but it does not take much to see that the conflict's roots were laid long before this century. Competence in technology investing does not confer expertise in geopolitics, yet the confidence remains.
The Enron story also illustrates what Stanford social psychologist Mary C. Murphy calls a «culture of genius» in her book Cultures of Growth. CEO Jeff Skilling built a company that recruited aggressively from Harvard, Stanford, and Wharton, where employees were expected to be exceptional and the company's identity became inseparable from the belief that it employed the smartest people in business. Murphy argues that such cultures often undermine learning. When talent is treated as innate, admitting a mistake or not knowing something becomes impossible. Employees stop questioning, exploring, or learning new things because geniuses, after all, already know everything.
Theo Baker described a similar environment at Stanford in his memoir How to Rule the World. A select few students deemed «high agency» gain access to a «Stanford within Stanford,» complete with mentorships, yacht party invitations, and venture capital to fund any idea. The notion that exceptional people are exempt from normal constraints is at least as old as Dostoevsky's Crime and Punishment. If you are a special, high-agency person, the rules that bind mere mortals do not apply, so why bother learning about the Zaporozhian Sich or the legal nuances of the Sherman and Clayton antitrust acts?
Artificial intelligence may be the ultimate high-agency technology. Like the trading and risk management techniques Enron pioneered in the 1990s, it is highly complex, little understood outside a select global elite, and incredibly powerful. Many believe it is advancing so quickly that only insiders can truly grasp what is happening. Yet the outlines of the Enron story are visible here too. While tech workers are blown away by how frontier models affect their own work, it is far from clear that similar gains will materialize across the rest of the economy. Research from the St. Louis Federal Reserve Bank shows that for most tasks, productivity gains are relatively meager. A recent Duke survey of CFOs found few see much improvement at all, and a survey of nearly 6,000 senior business executives across the U.S., U.K., Germany, and Australia reached similar conclusions.
The lesson is not that intelligence is worthless, but that it is domain-specific. Success in one arena can breed a dangerous certainty in others, and the higher the pedestal, the harder it becomes to admit ignorance. From Enron's trading floor to today's AI labs, the pattern repeats: the smartest people in the room are often the last to realize they are out of their depth.