AI systems improve themselves to learn and adapt better
The Last AI Built by Humans: Toward Genuine Recursive Self-Improvement
Machine LearningArtificial IntelligenceComputation and Language
Summary
AI systems often need help to get better at tasks and to improve how they learn over time. The paper introduces a step-by-step plan called recursive self-improvement (RSI), where AI not only learns but also improves how it learns and adapts from experience. The authors show challenges current models face and explain how RSI could work in different areas, like science or software development. They also link this idea to real-world applications and identify big hurdles to making true self-improving AI.
What this means in practice
- •For software engineering teams: Create AI tools that improve their own coding assistance skills and adapt strategies autonomously.
- •For robotics developers: Build embodied AI systems that can learn and refine their physical interactions with environments without constant human tuning.
A position paper. It proposes an approach and reports no results.
Authors
Yi Duan, Ying Liu, Zirui Tang, Haodong Chen, Jun Zhou, Yumou Liu, Bangrui Xu, Yukai Wu, Sidi Chen, Yuhan Zhou, Haoyu Wang, Xiaoyou Yu, Shaokun Han, Xuzhou Zhu, Le Zhou, Bolin Lu, Wei Zhou, Jiachen Liu, Nuozhou Fang, Jiaxin Tian, Ruoyu Chen, Yuxuan Li, Kai Zuo, Kaiyan Zhang, Jiantao Qiu, Conghui He, Guoliang Li, Bowen Zhou, Zhiyuan Liu, Zhoufutu Wen, Jihua Kang, Xuanhe Zhou, Fan Wu
Abstract
Recursive self-improvement (RSI) enables AI systems to turn experience and feedback into persistent changes that improve both their capabilities and the process of future improvement. We first use the Headroom-Closed Index (HCI) to reveal the problems of existing LLMs, then introduce the RSI concept and its development roadmap: from improvement-execution autonomy, improvement-strategy autonomy, experience-acquisition autonomy, and environment-adaptation autonomy, to recursive meta-improvement. Next we examine RSI across scenarios (e.g., scientific discovery, embodied intelligence, software engineering), highlighting their distinct requirements and development speeds. Drawing on diverse industry practices and preliminary empirical evidence, we connect RSI research with practical systems and identify key challenges to achieving genuine RSI.