AI Coaching for Accelerating Human Skill Development with Reinforcement Learning

2026-06-23

AI Summary

This study addresses a central challenge in human-AI collaborative learning: how to assist humans in task execution without fostering overreliance, thereby promoting long-term development of independent skills. The authors model the AI coach as engaging in a non-cooperative dynamic game with the human learner and propose an adaptive shared-control mechanism that integrates strategic scaffolding with timely withdrawal. For the first time, they formalize skill evolution as a causal consequence of coaching behaviors and optimize the coach’s policy using a probabilistic causal model. Evaluated in a first-person drone racing task with 33 human participants, their framework—combining reinforcement learning and causal inference—significantly outperforms existing approaches, effectively enhancing both learning outcomes and learners’ autonomy.

Abstract

AI copilots can substantially boost human performance through shared control, but excessive assistance can induce over-reliance and skill atrophy. This paper studies how an embodied AI agent can act as a coach that accelerates human motor-skill development. We argue that effective coaching requires strategic scaffolding and stepping back that are aligned with the learner's capability, allowing productive failures that drive learning. We formalize the interactive AI coaching process as a non-cooperative dynamic game in which the learner optimizes task performance while the coach targets the learner's independent competence. Building on this formalism, we develop a reinforcement learning framework combining adaptive shared control with probabilistic models of the coach's causal influence on skill evolution, enabling tractable training of coaching policies. A comprehensive user study (N=33) on first-person-view drone racing shows significant gains in human learning outcomes over state-of-the-art AI coaching baselines.

Problem

• AI coaching
• skill development
• over-reliance
• shared control
• motor-skill learning

Innovation

• reinforcement learning
• AI coaching
• adaptive shared control
• skill acquisition
• non-cooperative dynamic game