Towards AI as Colleagues: Multi-Agent System Improves Structured Professional Ideation

2025-10-27

AI Summary

Contemporary AI systems primarily function as task executors, lacking the capacity for genuine human-AI collaborative problem solving or creative co-creation. Method: This paper introduces the “AI Colleague” paradigm, instantiated as MultiColleagues—a multi-agent dialogue system that enables human-AI co-creation through agent-to-agent collaboration, shared intent modeling, and structured ideation frameworks (e.g., SCAMPER). It integrates social presence design principles and natural language interaction mechanisms to foster collective cognition and deepen user engagement. Contribution/Results: Empirical evaluation demonstrates that MultiColleagues significantly outperforms single-agent baselines across key creativity metrics: idea quality (+28.3%), novelty (+34.1%), and idea extensibility (+41.7%). These results validate the efficacy and scalability of multi-agent coordination in augmenting professional creative workflows through human-AI partnership.

Technology Category

• Humans and AI: Human-AI Collaboration / Human-AI Teaming
• Cognitive Modeling & Cognitive Systems: Computational Creativity
• Multiagent Systems: Teamwork

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Architectures and workflows that use LLMs for crowd work
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Most AI systems today are designed to manage tasks and execute predefined steps. This makes them effective for process coordination but limited in their ability to engage in joint problem-solving with humans or contribute new ideas. We introduce MultiColleagues, a multi-agent conversational system that shows how AI agents can act as colleagues by conversing with each other, sharing new ideas, and actively involving users in collaborative ideation. In a within-subjects study with 20 participants, we compared MultiColleagues to a single-agent baseline. Results show that MultiColleagues fostered stronger perceptions of social presence, produced ideas rated significantly higher in quality and novelty, and encouraged deeper elaboration. These findings demonstrate the potential of AI agents to move beyond process partners toward colleagues that share intent, strengthen group dynamics, and collaborate with humans to advance ideas.

Problem

• AI systems lack joint problem-solving with humans
• AI systems cannot contribute novel ideas effectively
• Single-agent systems limit collaborative ideation quality

Innovation

• Multi-agent conversational system for collaborative ideation
• AI agents converse and share new ideas
• Active user involvement in joint problem-solving