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

TL;DR

This work introduces AI Collaborator, a GPT-4–based platform for studying human-AI teaming by enabling researchers to craft customizable AI personas and a memory-driven interaction model. It integrates Slack for dialogue capture, a memory subsystem with dynamic response planning and a context-aware retrieval model, and a Big Five–informed persona customization pipeline, all under a researcher-facing UI with customization and analytics. The key innovations include a memory-augmented conversational agent with a formal retrieval strategy incorporating a Composite Score $ \text{Composite Score} = \alpha \times \text{Recency} + \beta \times \text{Relevance} + \gamma \times \text{Importance}$, decay $W(t) = e^{-\lambda t}$, and cosine-based similarity $ \text{Similarity}(M,M_i) = \frac{M \cdot M_i}{|M| |M_i|}$, plus a structured mechanism to translate personality traits into prompts and responses. By enabling controlled, scalable experiments across diverse team configurations, the platform offers a practical pathway to understanding and improving human-AI collaboration in educational and professional contexts.

Abstract

AI Collaborator, powered by OpenAI's GPT-4, is a groundbreaking tool designed for human-AI collaboration research. Its standout feature is the ability for researchers to create customized AI personas for diverse experimental setups using a user-friendly interface. This functionality is essential for simulating various interpersonal dynamics in team settings. AI Collaborator excels in mimicking different team behaviors, enabled by its advanced memory system and a sophisticated personality framework. Researchers can tailor AI personas along a spectrum from dominant to cooperative, enhancing the study of their impact on team processes. The tool's modular design facilitates integration with digital platforms like Slack, making it versatile for various research scenarios. AI Collaborator is thus a crucial resource for exploring human-AI team dynamics more profoundly.

Figures (2)

• Figure 1: Overview of the general workflow of AI Collaborator, visualizing the main components and how they interact.
• Figure 2: This prototype of the UI illustrates the adjustable expressions of the Big Five personality traits' subcomponents, with three levels of intensity: low, medium, and high. As depicted in this example, the interface presents a collaborative bot characterized by a highly dominant persona. This figure was adapted from the hierarchical model of facets related to teamwork, as detailed in driskell2006makes.