TECHNOLOGYBUSINESS INSIDER
What happens when AI starts gambling on soccer?
A startup tested AI models like ChatGPT, Gemini, and Claude by having them bet on World Cup matches using live Polymarket odds. Mistral led the field, followed by GPT 5.5 and DeepSeek V4, while Claude Opus 4.8 performed worst. The experiment evaluated AI judgment under uncertainty through virtual $10,000 bankroll wagers.
Mentioned
Related Signal
Adjacent reporting
- A soccer team bet against itself. This is the (good) future of prediction markets.
- AI models are terrible at betting on soccer—especially xAI Grok
- AI investor mania: China’s DeepSeek chases US$70 billion valuation in fresh round
- Blatant AI slop just won a 25k USD DeepMind Kaggle Grand Prize
- Hitting it big betting on startups in the AI era will be a lot harder, but Silicon Valley might not feel the worst of it