Episode Details

Fifty Seven Percent Wrong, Thirty Nine at Best: What Happened When Eighteen Chatbots Were Asked About Money, September 21, 2026 A British fintech firm put eighteen mainstream AI models through one hundred twenty one personal finance questions, asked five times each, and graded more than ten thousand answers. The models were wrong fifty seven percent of the time, eighty eight percent on complex multi step questions, and the single best performer still failed on thirty nine percent. Chris and Laura dig into the methodology, the costly real world errors it surfaced, the free versus paid gap, and why personal finance is uniquely hostile territory for a language model. Hosted by Chris and Laura. The DX Today Podcast brings you daily deep dives into the most consequential stories in the AI ecosystem. #ArtificialIntelligence #AISafety #PersonalFinance #FinTech #AIRegulation