AI Pioneer Who Built First Hedge Fund Won't Trust ChatGPT With His Money
Vasant Dhar helped bring AI to Wall Street in 1994. His skepticism of today's tools carries weight investors should hear.
Vasant Dhar occupies a rare position in the debate over artificial intelligence and investing: he was there at the beginning. In 1994, Dhar helped pioneer the use of machine learning in financial markets, building one of the earliest AI-driven hedge funds at a time when the technology was barely understood outside academic circles. That foundational credential makes his current reservations about tools like ChatGPT more than casual skepticism — it is informed caution from someone who has seen both the promise and the limits of algorithmic decision-making up close.
Dhar's core concern appears to center on the distinction between AI systems designed and validated specifically for financial prediction versus large language models built for general-purpose conversation. The latter, however impressive in generating text, were not architected to handle the probabilistic, adversarial, and constantly shifting nature of financial markets. Trusting a chatbot with capital allocation, in his view, conflates fluency with competence — a confusion that could prove costly for retail investors drawn in by the technology's surface-level confidence.
Read more Strategy Raises $544M in Stock Sales, Buys Back Preferred Shares →
His perspective offers a useful corrective to the breathless enthusiasm surrounding AI-powered investing tools that have proliferated since ChatGPT's public launch. Financial markets are among the most competitive information environments on earth; edges erode quickly, and models that cannot adapt to regime changes or account for their own market impact are structurally limited. Dhar's three decades of experience suggest that rigorous, purpose-built systems — not general AI assistants — are what serious quantitative investing actually requires.
For everyday investors, the practical takeaway is a call for discernment. The fact that a tool can answer questions about a stock does not mean it can reliably predict that stock's behavior. As AI continues to be marketed as a financial advisor replacement, Dhar's skepticism serves as a reminder that provenance, design intent, and empirical validation matter far more than raw computational power or conversational polish.
Continue reading at MarketWatch.com
