The Pioneer Who Brought AI to Wall Street Won't Trust ChatGPT With His Money
Vasant Dhar helped build one of Wall Street's first AI-driven hedge funds in 1994 and has pointed warnings for investors seduced by today's AI tools.
Vasant Dhar was doing artificial intelligence on Wall Street before most of today's fintech evangelists had opened their first brokerage account. In 1994, he helped construct one of the earliest AI-driven hedge funds — a milestone that places him in rare company when it comes to understanding both the promise and the hard limits of machine-driven investing. Three decades later, he is not impressed enough by tools like ChatGPT to hand them any real authority over his own money.
That skepticism from a genuine pioneer carries analytical weight that casual AI enthusiasm typically lacks. Dhar's concern is not that modern large language models are useless — it is that they are being misapplied. Financial decision-making demands precise, probabilistic reasoning grounded in structured data and rigorous back-testing. Generative AI, by contrast, excels at pattern-matching across language, which makes it a capable research assistant but a potentially dangerous portfolio manager. The gap between those two roles is enormous, and conflating them is the kind of mistake that tends to be expensive.
Read more Binance Expands Into Gold and Silver Options Trading →
For retail investors, the warning is especially pertinent. The accessibility of tools like ChatGPT creates an illusion of expertise — a fluent, confident voice that can discuss asset allocation or earnings forecasts without actually bearing accountability for the outcomes. Dhar's three decades of experience suggest that the discipline separating useful quantitative AI from noise is not the sophistication of the model but the quality of the data pipeline, the specificity of the objective function, and the humility to stress-test every assumption.
The broader lesson Dhar offers the market is one of institutional maturity. AI in finance is not new, and the firms that have used it most effectively have done so quietly, methodically, and with deep respect for how quickly models degrade when market regimes shift. The current wave of generative AI hype risks short-circuiting that hard-won discipline, tempting both individual investors and institutions to over-rely on tools optimized for conversation rather than capital allocation.
Continue reading at MarketWatch.com
