EVENT
The DBOT Test
The 2024 NYU experiment in which Vasant Dhar built an AI trained on Damodaran's entire output — a bot that imitated his voice but failed to replicate his judgment, becoming the empirical demonstration of where the AI moat actually sits.
DBOT is the AI entity built by NYU machine learning professor Vasant Dhar and his team in 2024, trained on every blog post, lecture, valuation, and book Damodaran has published over four decades. It can value any publicly traded company, produce comprehensive reports in Damodaran's prose style, and follow Damodaran's stated methodology. When tested against Damodaran's own published valuations, DBOT's outputs came within plus-or-minus fifty percent of market value — a range its creators considered encouraging. But DBOT has a specific limitation that the research team itself documented: "produced reports in the linguistic style of Damodaran, but failed to capture his analysis and thus lacked credible valuations." The bot replicated voice; it did not replicate framing. This gap — between voice and judgment — is the empirical demonstration of where AI-resistant moats actually sit, and it became the founding case study for Damodaran's 2024 "Beat Your Bot" essay.
