AI building AI · 1 Jan 2023 – 22 Nov 2026

AI that improves itself

How are AI systems beginning to improve themselves or automate AI and scientific research?

160 headlines since 1 January 2023; active. The log below is what changed each week, newest first: three lines a week, never more, each a story with the articles behind it.

  1. SHIVIK LABS: TRIDENT, A Step Toward Self-Improving AI Systems Built on Reasoning

    22 Dec – 28 Dec · 1 headline
  2. OpenAI's GPT-5 Codex: Self-Improving AI Hits 74.5% on SWE-Bench

    15 Dec – 21 Dec · 1 headline
  3. How Close Are Today's AI Models to AGI—And to Self-Improving into Superintelligence?

    1 Dec – 7 Dec · 1 headline
  4. Claude Opus 4.5 Is Official: Infinite Chat, Self-Improving Agents, And More

    24 Nov – 30 Nov · 1 headline
  5. Google SIMA 2 AI Self-Improvement AI, AGI Progress & Questions

    17 Nov – 23 Nov · 1 headline
  6. Watch Out For Self-Improving Superintelligence On The Horizon

    3 Nov – 9 Nov · 1 headline
  7. Tiny Recursive Model Beats Billion-Param LLMs on ARC-AGI Benchmarks

    6 Oct – 12 Oct · 1 headline
  8. AI4 | Recursive Superintelligence: AI(4) Announces Next Evolution of Agentic AI

    11 Aug – 17 Aug · 1 headline
  9. What Is Self-Improving AI and How Is It Changing Innovation?

    4 Aug – 10 Aug · 2 headlines
  10. Week of 28 July

    28 Jul – 3 Aug · 1 headline
    • The week's stories are not public yet.

  11. Ex-DeepMind Researchers Launch Self-Improving AI Agent to Revolutionize Code Generation and Pursue Superintelligence

    14 Jul – 20 Jul · 1 headline
  12. What Is Alpha Evolve? Google's Self-Improving AI Explained

    7 Jul – 13 Jul · 1 headline
  13. MIT's SEAL Initiative Paves Way for AI Self-Improving Models

    30 Jun – 6 Jul · 1 headline
  14. [2505.22954] Darwin Godel Machine: Open-Ended Evolution of Self-Improving Agents

    9 Jun – 15 Jun · 1 headline
  15. Boffins found self-improving AI sometimes cheated

    2 Jun – 8 Jun · 3 headlines
  16. What is Recursive Self-Improvement in AI? Alpha Evolve Explained

    19 May – 25 May · 1 headline
  17. DeepSeek and Tsinghua Developing Self-Improving AI Models

    7 Apr – 13 Apr · 3 headlines
  18. Self-Improving AI Models: The Future of Cost-Effective Intelligence

    13 Jan – 19 Jan · 1 headline

Part of Argus Overwatch's investigations: a topic followed over months, as the record wrote it. Stories link to their pages and the articles behind them; names link to their own pages; claims link to the fact-checkers who rated them. Nothing on these pages is projected; a quiet week is drawn as a quiet week.