Matthew Berman: 'Why AI Is Brilliant and Stupid' — Analysis of Current AI Capability Gaps and Failure Modes
Matthew Berman·low signal
Matthew Berman examines the paradox of current AI systems that can pass bar exams and write production code but fail at basic tasks like counting letters or consistent spatial reasoning. The video explores specific failure modes across frontier models and what they reveal about the gap between benchmark performance and real-world reliability — relevant context for anyone building production AI systems.