🤖 AI Fails Premier League Betting Test

Eight popular AI models were each given €115,000, bookmaker odds and historical data from the 2023/24 Premier League season to test whether they could profit from sports betting. None delivered a positive return. Claude performed best at roughly -11%, while one OpenAI model lost about 13%; Grok often burned through the full bankroll. 📉🤖 The study by General Reasoning aimed to mirror real betting behaviour, including stake sizing and strategy changes over time. Even with possible prior knowledge of parts of the season, the models still lost money — a sharp reminder that AI hype has not yet translated into reliable betting edge. 🎯⚽ Source: General Reasoning, via highroller_channel Telegram.