Dr. He Wen, PI of AI Safety Lab, Indiana State University

Safely Apply AI in Engineering

• AI risks originate from intrinsic system flaws, but manifest through human–AI interaction.
• Human–AI conflict is the most typical manifestation of such risks in human-AI interaction.
• I am the first to propose and mathematically model “human–AI conflict”.
• I study AI from the perspective of safety engineering, rather than computer science.
• What actually happens when humans and AI jointly control a safety-critical system?