How does artificial intelligence improve human decision‐making? Evidence from the AI‐powered Go program

Abstract Research Summary We study how humans learn from artificial intelligence (AI), leveraging an introduction of an AI‐powered Go program (APG) that unexpectedly outperformed the best professional player. We compare the move quality of professional players to APG's superior solutions around its public release. Our analysis of 749,190 moves demonstrates significant improvements in players' move quality, especially in the early stages of the game where uncertainty is highest. This improvement was accompanied by a higher alignment with AI's suggestions and a decreased number and magnitude of errors. Young players show greater improvement, suggesting potential inequality in learning from AI. Further, while players of all skill levels benefit, less skilled players gain higher marginal benefits. These findings have implications for managers seeking to adopt and utilize AI in their organizations. Managerial Abstract We examine how professionals can learn from artificial intelligence (AI) by studying an AI‐powered Go program (APG) that outperformed the best professional player. By analyzing 749,190 moves, we find that players' move quality improved significantly, closely aligning with the AI's recommendations. The number and magnitude of errors also decreased. This learning effect was particularly strong early in the game where decisions are more uncertain. Young players showed greater effect, suggesting that learning from AI may vary by age. While players of all skill levels benefited, those with less skill saw the greatest improvement. These findings highlight the instructional role of AI and offer guidance on how to effectively integrate AI into organizations to enhance worker performance across different age groups and skill levels.