Co-Creative Learning via Metropolis-Hastings Interaction between Humans and AI

2025-06-18

AI Summary

This study addresses the fundamental challenge of human-AI collaborative creative learning, moving beyond traditional unidirectional knowledge transfer to investigate how humans and AI jointly construct and share external representations—i.e., emergent symbolic systems—under partial observability. Method: We propose a decentralized Bayesian inference framework based on Metropolis-Hastings (MH) sampling, integrated with a joint-attention naming game (JA-NG) and online human-AI interaction experiments. Contribution/Results: For the first time, we empirically demonstrate that this mechanism enables spontaneous co-emergence of a shared symbol system within the human-AI dyad. Experiments show that MH-based AI agents significantly improve human classification accuracy and accelerate symbolic convergence; human adoption behavior aligns closely with theoretical acceptance probabilities (p < 0.001). Our work establishes “co-learning with humans” as a foundational paradigm for symbiotic AI—distinct from “learning from humans”—and provides a theoretically grounded, empirically testable framework for modeling cross-modal collaborative cognition.

Technology Category

• Humans and AI: Learning Human Values and Preferences
• Cognitive Modeling & Cognitive Systems: Symbolic Representations
• Machine Learning: Neuro-Symbolic Learning

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: Research challenges in human and human-AI computation
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

We propose co-creative learning as a novel paradigm where humans and AI, i.e., biological and artificial agents, mutually integrate their partial perceptual information and knowledge to construct shared external representations, a process we interpret as symbol emergence. Unlike traditional AI teaching based on unilateral knowledge transfer, this addresses the challenge of integrating information from inherently different modalities. We empirically test this framework using a human-AI interaction model based on the Metropolis-Hastings naming game (MHNG), a decentralized Bayesian inference mechanism. In an online experiment, 69 participants played a joint attention naming game (JA-NG) with one of three computer agent types (MH-based, always-accept, or always-reject) under partial observability. Results show that human-AI pairs with an MH-based agent significantly improved categorization accuracy through interaction and achieved stronger convergence toward a shared sign system. Furthermore, human acceptance behavior aligned closely with the MH-derived acceptance probability. These findings provide the first empirical evidence for co-creative learning emerging in human-AI dyads via MHNG-based interaction. This suggests a promising path toward symbiotic AI systems that learn with humans, rather than from them, by dynamically aligning perceptual experiences, opening a new venue for symbiotic AI alignment.

Problem

• Integrating partial perceptual information between humans and AI
• Addressing modality differences in human-AI knowledge transfer
• Establishing shared representations via decentralized Bayesian inference

Innovation

• Metropolis-Hastings naming game for human-AI interaction
• Decentralized Bayesian inference for shared representations
• Dynamic alignment of perceptual experiences