Academic Lecture: Reliability of Human-Intelligence Team Formation

Human-AI Teaming Reliability

Liu Peng, Zhejiang University

January 9, 2026 (Friday) 15:00

From the perspective of reliability analysis, this talk explores whether Human-AI Machine Teaming (HMT) can achieve human-machine synergy and complementarity. A literature review and reliability analysis of HMT applications in a typical safety-critical domain (i.e., healthcare) were conducted. The study found that in most cases, HMT did not achieve full synergy or high complementarity; however, HMT reliability outperformed that of doctors working alone, indicating the significant potential of human-AI teaming in improving healthcare quality. In a few cases, human-machine conflicts occurred, where HMT reliability was lower than that of doctors working alone. Two categories of factors influence HMT complementarity: (1) teaming mode—synchronous (doctors see diagnostic tasks and AI results simultaneously) vs. sequential (doctors perform independent diagnosis first, then integrate AI results); (2) doctor experience—HMT more effectively improves reliability for less experienced doctors. In addition, a nonlinear quantitative model for HMT reliability was developed.

Liu Peng, Researcher under the first category of Zhejiang University’s “New Hundred Talents Program”, Doctoral Supervisor; PhD in Human Factors Engineering from Tsinghua University; Visiting Scholar at the Department of Experimental Psychology and Faculty of Philosophy, University of Oxford; Visiting Professor at the National University of Singapore. Engages in human factors engineering and psychological research on autonomous vehicles and AI systems. Published 60 SCI/SSCI journal papers as first or corresponding author, including 4 ESI Hot Papers and Highly Cited Papers. Led 3 National Natural Science Foundation of China projects and 1 key project sub-task. Work published in journals such as Engineering, The Innovation, Risk Analysis, Journal of Experimental Psychology: Applied, Cognition, npj Artificial Intelligence, and reported multiple times by international media such as ScienceDaily. Received the IEA/Kingfar Early Career Researcher Award from the International Ergonomics Association in 2024; won 5 international conference best/outstanding paper awards and 1 best paper nomination. Continuously listed among the World’s Top 2% Scientists from 2021 to 2025.

Prepared by: Ren Luyang

Typeset by: Chen Shifa

First Review and Proofreading: Ren Luyang

Second Review and Proofreading: Ma Jiang

Third Review and Proofreading: Zheng Chun