Reddit user Nathan Smith handed $100 and full trading responsibility to GPT-4o, OpenAI’s latest AI model. In just one month, the AI managed to outperform major market indices — including the S&P 500. But experts warn: don’t ditch human judgment just yet.
Gpt-4O Turns $100 Into $125 in Four Weeks
Nathan Smith’s experiment started with a simple idea: could a publicly available AI model beat the stock market with a small budget and no prior financial expertise? He gave GPT-4o, OpenAI’s most recent release, control over a $100 portfolio and asked it to pick small-cap stocks over a six-month period. At the one-month mark, he reported a 25% gain — far outpacing the S&P 500, which rose less than 3% during the same period.
Smith shared a detailed breakdown of the results on Reddit, along with visual performance graphs and a GitHub page outlining his methodology. The AI model selected stocks daily based on recent performance trends and sentiment data. Smith then manually executed the trades and applied a simple stop-loss rule to minimize downside risk — selling any stock that dropped below a preset threshold.
According to Smith, GPT-4o consistently recommended stocks that had short-term upside potential, often tapping into biotech and emerging tech companies. “It was evident early on that the model wasn’t just guessing,” he wrote. “The picks made sense — and the results spoke for themselves.”
AI Outperformed Traditional Indices — but Under Strict Conditions
While the early performance looks strong on paper, it’s important to put the results in context. The portfolio size was extremely small, and trades were executed manually with heavy human oversight. Smith noted that although GPT-4o generated recommendations, he was the one responsible for applying trading rules and interpreting unclear results.
This isn’t the first time AI models have been used to pick stocks. A 2024 study from the University of Duisburg-Essen in Germany found that certain OpenAI models could spot patterns leading to short-term gains. But researchers stressed that success depended on careful prompt design and limited trade volume. In larger-scale simulations, the performance often leveled out or declined.
Finance professor Alejandro Lopez-Lira from the University of Florida echoed this caution in an interview with Morningstar. “Our paper-based results are far more optimistic than real-world outcomes with normal portfolio sizes,” he said. AI models, he added, struggle with long-term forecasting, especially when market conditions shift rapidly.
If AI Investing Works, What Happens When Everyone Uses It?
One overlooked consequence of successful AI investing is its potential to disrupt the very market it tries to profit from. If too many people use the same algorithm to pick stocks, the predictive edge disappears. Prices adjust faster, and the strategies that once worked may lose their effectiveness.
The website Futurism picked up Smith’s story, noting that even if GPT-4o continues to beat the market for now, widespread adoption could nullify its advantages. AI’s strength lies in spotting inefficiencies — but if those inefficiencies vanish, so does the opportunity.
This echoes broader concerns raised by experts in algorithmic trading, where over-reliance on similar strategies can create feedback loops and market instability. Smith seems aware of this. In his GitHub documentation, he emphasized that this is “just an experiment,” not a reliable blueprint for making money with AI.
AI Still Needs a Human in the Loop — at Least for Now
The experiment also highlights how far we are from fully autonomous investing. Despite GPT-4o’s role, Smith remained an active participant: feeding it daily data, interpreting output, and enforcing basic trading rules.
While the time commitment was low — “just a few minutes a day,” according to Smith — the process showed that human involvement is still essential, especially when dealing with volatile markets like small-cap equities. Even small misinterpretations or model hallucinations can lead to costly mistakes.