Abstract:We explore the potential for productive team-based collaboration between humans and Artificial Intelligence (AI) by presenting and conducting initial tests with a general framework that enables multiple human and AI agents to work together as peers. ChatCollab's novel architecture allows agents - human or AI - to join collaborations in any role, autonomously engage in tasks and communication within Slack, and remain agnostic to whether their collaborators are human or AI. Using software engineering as a case study, we find that our AI agents successfully identify their roles and responsibilities, coordinate with other agents, and await requested inputs or deliverables before proceeding. In relation to three prior multi-agent AI systems for software development, we find ChatCollab AI agents produce comparable or better software in an interactive game development task. We also propose an automated method for analyzing collaboration dynamics that effectively identifies behavioral characteristics of agents with distinct roles, allowing us to quantitatively compare collaboration dynamics in a range of experimental conditions. For example, in comparing ChatCollab AI agents, we find that an AI CEO agent generally provides suggestions 2-4 times more often than an AI product manager or AI developer, suggesting agents within ChatCollab can meaningfully adopt differentiated collaborative roles. Our code and data can be found at: <a class="link-external link-https" href="https://github.com/ChatCollab" rel="external noopener nofollow">this https URL</a>.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to explore the potential for effective collaboration between humans and artificial intelligence (AI) agents in teams. Specifically, the paper presents a general - purpose framework named ChatCollab, which supports multiple humans and AI agents to work together as equals. ChatCollab is designed to: 1. **Role Identification and Responsibility Allocation**: AI agents can identify their roles and responsibilities in the team and participate in tasks and communication autonomously based on these roles. 2. **Coordination and Communication**: AI agents can carry out tasks and communicate autonomously in Slack without concerning whether their collaborators are humans or AI. 3. **Adapting Behaviors to Different Roles**: Through different prompts, AI agents can exhibit behavioral characteristics that match their roles. For example, an AI CEO offers advice 2 - 4 times more frequently than an AI product manager or an AI developer. 4. **Practical Applications in Software Development**: Taking software engineering as an example, the performance of AI agents in actual projects was studied. It was found that the software quality produced by ChatCollab's AI agents in interactive game development tasks is comparable to or better than that of other multi - AI agent systems. 5. **Collaboration Dynamic Analysis**: An automated method for analyzing collaboration dynamics is proposed. This method can effectively identify the behavioral characteristics of agents with different roles, thereby quantitatively comparing the collaboration dynamics under different experimental conditions. Through these goals, the paper aims to promote the collaboration between humans and AI agents, making it not just a simple auxiliary tool, but a true partner capable of undertaking diverse roles and complex tasks. This provides new perspectives and possibilities for future human - AI collaboration.

ChatCollab: Exploring Collaboration Between Humans and AI Agents in Software Teams

Group Chat Ecology in Enterprise Instant Messaging: How Employees Collaborate Through Multi-User Chat Channels on Slack

Multi-Agent Software Development through Cross-Team Collaboration

Human AI Collaboration in Software Engineering: Lessons Learned from a Hands On Workshop

The AI Collaborator: Bridging Human-AI Interaction in Educational and Professional Settings

A System for Human-AI collaboration for Online Customer Support

ChatDev: Communicative Agents for Software Development

Human-AI Teamwork Interface Design Using Patterns of Interactions

AgentGroupChat: An Interactive Group Chat Simulacra For Better Eliciting Emergent Behavior

Human-AI Collaboration Enables More Empathic Conversations in Text-based Peer-to-Peer Mental Health Support

Human-AI Collaboration in a Cooperative Game Setting

InterAct: Exploring the Potentials of ChatGPT as a Cooperative Agent

How Software Developers Mitigate Collaboration Friction with Chatbots

Teaming up with information agents

CAMEL: Communicative Agents for "Mind" Exploration of Large Language Model Society

AI Chatbots as Multi-Role Pedagogical Agents: Transforming Engagement in CS Education

ChatISA: A Prompt-Engineered Chatbot for Coding, Project Management, Interview and Exam Preparation Activities

From Human-Human Collaboration to Human-AI Collaboration

Designs for Enabling Collaboration in Human-Machine Teaming via Interactive and Explainable Systems

Conversational Swarms of Humans and AI Agents enable Hybrid Collaborative Decision-making

Roots and Requirements for Collaborative AIs