![[2504.20903] Modeling AI-Human Collaboration as a Multi-Agent Adaptation](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2504.20903] Modeling AI-Human Collaboration as a Multi-Agent Adaptation
Summary
This paper explores AI-human collaboration through agent-based simulations, revealing how distinct decision-making heuristics impact performance in modular and sequenced tasks.
Why It Matters
Understanding the dynamics of AI-human collaboration is crucial as organizations increasingly integrate AI into decision-making processes. This research provides insights into optimizing collaboration based on task architecture, which can enhance productivity and decision quality across various sectors.
Key Takeaways
- AI typically substitutes for humans in modular tasks, but complementarities can emerge under certain conditions.
- In sequenced tasks, human-led searches followed by AI refinement yield the best performance, challenging traditional AI-first approaches.
- Memory-less random AI can improve outcomes for low-capability humans by helping them escape local optima.
- Task architecture, including division of labor and interdependence, is critical for effective AI-human collaboration.
- The findings suggest a need to rethink AI integration strategies in organizational settings.
Computer Science > Multiagent Systems arXiv:2504.20903 (cs) [Submitted on 29 Apr 2025 (v1), last revised 15 Feb 2026 (this version, v3)] Title:Modeling AI-Human Collaboration as a Multi-Agent Adaptation Authors:Prothit Sen, Sai Mihir Jakkaraju View a PDF of the paper titled Modeling AI-Human Collaboration as a Multi-Agent Adaptation, by Prothit Sen and 1 other authors View PDF Abstract:We formalize AI-human collaboration through an agent-based simulation that distinguishes optimization-based AI search from satisficing-based human adaptation. Using an NK model, we examine how these distinct decision heuristics interact across modular and sequenced task structures. For modular tasks, AI typically substitutes for humans, yet complementarities emerge when AI explores a moderately broad search space and human task complexity remains low. In sequenced tasks, we uncover a counterintuitive result: when a high-performing human initiates search and AI subsequently refines it, joint performance is maximized, contradicting the dominant AI-first design principle. Conversely, when AI leads and human satisficing follows, complementarities attenuate as task interdependence increases. We further show that memory-less random AI, despite lacking structured adaptation, can improve outcomes when augmenting low-capability humans by enabling escape from local optima. Collectively, our findings reveal that effective AI-human collaboration depends less on industry context and more on task architec...
