Humans set goals for AI agents, but the agent determines how to achieve the goal, proactively performing supportive tasks with little or no human intervention. AI agents can react to changes in their environment (more ably if environments are well defined), responding dynamically and adapting in ways that enable them to effectively complete often complex tasks more effectively than their human counterparts. Unlike traditional AI assistants, AI agents can interact with and make use of external datasets, applications, even other agents and things in the physical world. Compared to a basic query.