Machine Learning 29 November 2023The new method uses crowdsourced feedback to help train robots To teach an AI agent a new task, such as how to open a kitchen cabinet, researchers often use reinforcement…