Proximal State Nudging: Reducing Skill Atrophy from AI Assistance

TL;DR

Proximal State Nudging (PSN) is a novel shared autonomy algorithm that enhances skill retention and safety by guiding users toward learnable states, outperforming existing methods in simulations and human studies.

Contribution

The paper introduces PSN, the first shared autonomy planner designed to improve skill development and safety, validated through simulations and human experiments.

Findings

PSN outperforms existing shared autonomy baselines in the LunarLander environment.

In human driving tasks, PSN yields up to 7x larger skill gains than standard methods.

PSN reduces collisions by 50% compared to unassisted self-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.