The study and design of how people use computers, aimed at making technology usable, useful and safe for the people using it. It covers everything from button layout to how far people trust an automated suggestion. With AI, it asks new questions: how a system should show uncertainty, explain itself, and let people correct it.
Technical view
Human-computer Interaction (‘HCI’) combines cognitive psychology, design and computer science, with methods including user research, usability testing, controlled experiments and field studies. For AI systems the questions include how to support appropriate reliance rather than blind trust or blanket rejection, how to design conversational and agent interfaces whose behaviour varies from one run to the next, and how to make outputs inspectable and correctable. Findings from automation research, such as complacency and loss of skill, apply directly. Accessibility and evaluation with real users, rather than benchmarks alone, are central concerns.
Related
Part of AI Terminology Index, section: Adjacent Fields.