Darwin Godel Machine: Open-Ended Evolution of Self-Improving Agents

2025-05-29

AI Summary

This work addresses the fundamental limitation of current AI systems—static architectures and inability to autonomously sustain long-term evolution. We propose the first open-ended evolutionary framework integrating Darwinian variation with Gödelian self-modification. Grounded in foundation models, the framework enables autonomous agent mutation, sandboxed empirical validation, dynamic maintenance of a diversity archive, long-context co-evolution, and peer-review-based selection—supporting unbounded innovation steps and automatic discovery of “stepping-stone” solutions. Evaluated on SWE-bench and Polyglot benchmarks, our approach achieves accuracy improvements from 20.0% to 50.0% and from 14.2% to 30.7%, respectively, significantly outperforming baselines lacking self-improvement or constrained exploration. The framework establishes a scalable paradigm for developing safe, robust, and perpetually self-evolving artificial general intelligence systems.

Technology Category

• Machine Learning: Evolutionary Learning
• Search and Optimization: Evolutionary Computation
• Philosophy and Ethics of AI: Artificial General Intelligence

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Agentic search
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

Today's AI systems have human-designed, fixed architectures and cannot autonomously and continuously improve themselves. The advance of AI could itself be automated. If done safely, that would accelerate AI development and allow us to reap its benefits much sooner. Meta-learning can automate the discovery of novel algorithms, but is limited by first-order improvements and the human design of a suitable search space. The G""odel machine proposed a theoretical alternative: a self-improving AI that repeatedly modifies itself in a provably beneficial manner. Unfortunately, proving that most changes are net beneficial is impossible in practice. We introduce the Darwin G""odel Machine (DGM), a self-improving system that iteratively modifies its own code (thereby also improving its ability to modify its own codebase) and empirically validates each change using coding benchmarks. Inspired by Darwinian evolution and open-endedness research, the DGM maintains an archive of generated coding agents. It grows the archive by sampling an agent from it and using a foundation model to create a new, interesting, version of the sampled agent. This open-ended exploration forms a growing tree of diverse, high-quality agents and allows the parallel exploration of many different paths through the search space. Empirically, the DGM automatically improves its coding capabilities (e.g., better code editing tools, long-context window management, peer-review mechanisms), increasing performance on SWE-bench from 20.0% to 50.0%, and on Polyglot from 14.2% to 30.7%. Furthermore, the DGM significantly outperforms baselines without self-improvement or open-ended exploration. All experiments were done with safety precautions (e.g., sandboxing, human oversight). The DGM is a significant step toward self-improving AI, capable of gathering its own stepping stones along paths that unfold into endless innovation.

Problem

• Developing self-improving AI with autonomous continuous architecture evolution
• Overcoming limitations of meta-learning and human-designed search spaces
• Ensuring safe empirical validation of code modifications in AI systems

Innovation

• Self-improving AI via iterative code modifications
• Darwinian evolution-inspired open-ended agent archive
• Foundation model-aided parallel exploration of agents