Collaborative Intelligence in Sequential Experiments: A Human-in-the-Loop Framework for Drug Discovery

Drug discovery is a complex process that involves sequentially screening and examining a vast array of molecules to identify those with the target properties. This process faces challenges because of the vast search space, the rarity of target molecules, and constraints imposed by limited data and experimental budgets. To overcome these challenges, we propose a human-centered human–algorithm collaboration framework. Notably, both the algorithm and humans have substantial knowledge gaps. The algorithm proposes, and human experts retain decision rights to approve, revise, or override. Our design artifact leverages dual-process theory for attention management, surfaces meta-knowledge as a shared state for coordination, and optimizes a joint team objective. In real-world drug discovery tasks, the human–AI team consistently outperforms all baselines, including human-only and AI-only methods. This demonstrates complementary performance. These findings illustrate that the optimal use of AI in complex decision-making environments is to augment, rather than replace, human expertise. Managers should balance AI integration with the cultivation of deep domain expertise, ensuring that technology enhances decisions without eroding uniquely human strengths. For practice and policy: preserve human agency with clear decision rights, make model uncertainty explicit, and use AI to route scarce expert attention to the most informative experiments.