When Do Humans Heed AI Agents' Advice? When Should They?

OBJECTIVE: We manipulate the presence, skill, and display of artificial intelligence (AI) recommendations in a strategy game to measure their effect on users' performance. BACKGROUND: Many applications of AI require humans and AI agents to make decisions collaboratively. Success depends on how appropriately humans rely on the AI agent. We demonstrate an evaluation method for a platform that uses neural network agents of varying skill levels for the simple strategic game of Connect Four. METHODS: We report results from a 2 × 3 between-subjects factorial experiment that varies the format of AI recommendations (categorical or probabilistic) and the AI agent's amount of training (low, medium, or high). On each round of 10 games, participants proposed a move, saw the AI agent's recommendations, and then moved. RESULTS: Participants' performance improved with a highly skilled agent, but quickly plateaued, as they relied uncritically on the agent. Participants relied too little on lower skilled agents. The display format had no effect on users' skill or choices. CONCLUSIONS: The value of these AI agents depended on their skill level and users' ability to extract lessons from their advice. APPLICATION: Organizations employing AI decision support systems must consider behavioral aspects of the human-agent team. We demonstrate an approach to evaluating competing designs and assessing their performance.

Communicating scientific uncertainty.

All science has uncertainty. Unless that uncertainty is communicated effectively, decision makers may put too much or too little faith in it. The information that needs to be communicated depends on the decisions that people face. Are they (i) looking for a signal (e.g., whether to evacuate before a hurricane), (ii) choosing among fixed options (e.g., which medical treatment is best), or (iii) learning to create options (e.g., how to regulate nanotechnology)? We examine these three classes of decisions in terms of how to characterize, assess, and convey the uncertainties relevant to each. We then offer a protocol for summarizing the many possible sources of uncertainty in standard terms, designed to impose a minimal burden on scientists, while gradually educating those whose decisions depend on their work. Its goals are better decisions, better science, and better support for science.