Giving AI a classic psychological test reveals an inherent weakness in LLM decision-making abilities. Suketu Patel and ...