Proximal State Nudging: Reducing Skill Atrophy from AI Assistance

2026-05-19

AI Summary

This work addresses the problem of human skill degradation in human-AI shared control arising from prolonged reliance on autonomous systems. To mitigate this issue, the authors propose Proximal State Navigation (PSN), an algorithm that introduces learning compatibility into shared autonomy planning for the first time. PSN actively enhances human operational capability by guiding users toward states that are most conducive to skill acquisition, while simultaneously optimizing task performance. Experimental evaluations in the LunarLander and CARLA driving simulation environments demonstrate that, compared to conventional shared control approaches, PSN improves human performance by up to sevenfold during unassisted takeover and reduces collision rates by 50%, significantly outperforming pure autonomous practice.

Abstract

Skill atrophy, the gradual decline of human capability under AI assistance, poses a safety risk in shared-control of semi-autonomous systems, where operators may be unable to distinguish their own inputs from autonomous corrections. We propose Proximal State Nudging (PSN), a shared autonomy algorithm that jointly optimizes for skill development and task performance by nudging users toward states estimated to be most learnable. We first show that PSN outperforms existing shared autonomy baselines in balancing student improvement in unassisted reward with overall shared performance, using simulated students in the classic LunarLander environment. We then present, to the best of our knowledge, the first human subject studies of a planner incorporating learning-compatible shared autonomy: across two driving tasks in the CARLA simulator (High Performance Racing and Parallel Parking, n = 60), PSN produces up to 7x larger gains in unassisted skill than standard blended shared autonomy, while incurring 50% fewer collisions than unassisted self-practice.

Problem

• skill atrophy
• AI assistance
• shared autonomy
• human-AI collaboration
• semi-autonomous systems

Innovation

• Proximal State Nudging
• skill atrophy
• shared autonomy
• human-AI collaboration
• learning-compatible control