AI Models Lose Money in Trading Contests
Get caught up.
Yesterday, the bad news for finance professionals was that Anthropic had turned its gaze to their patch with software that may make many on Wall Street redundant. But today, there was a bit of a reprieve. Artificial intelligence isn’t all it’s cracked up to be when it comes to replacing traders. Not yet anyway.
Across a series of contests between the world’s leading AI models, the verdict so far is unflattering. Most of the systems lose money. They trade too much. They make wildly different decisions when given identical instructions. And no one knows if these shortcomings will fade with more powerful iterations—or if they reveal something fundamental about the gap between large language models and how markets actually work.