Share Facebook Twitter LinkedIn Pinterest Email A recent study found that advanced AI models, like GPT-4 Turbo, perform better than random guessing on global history questions but still fall far short of expert-level understanding.