What Is RSI? Recursive Self-Improvement in AI, Explained
Recursive self-improvement (RSI) means AI helping build better AI. Here is what it actually involves, how close labs are, and why it matters for polic
Tag archive
Recursive self-improvement (RSI) means AI helping build better AI. Here is what it actually involves, how close labs are, and why it matters for polic
SwarmRSI Hits the Spotlight: 2026’s First Open‑Ended Recursive Self‑Improvement...
Memento 3 Puts Recursive Self‑Improvement on the Open‑Source Shelf “The moment an AI can...
Weco says its AIDE² outer loop produced seven accepted harness rewrites over eight days and beat a human-engineered baseline under fixed budgets, while its test for compounding self-improvement remained inconclusive.

Anthropic's new R&D Automation Index shows AI now leads a quarter of the company's own AI research, collaborates on more than 90% of it, and runs as 30,000 agents at once, watched mostly by other machines. Fully autonomous work: still zero. The compa
Dream-RSI uses replayed discovery histories to improve the exploration policy around a fixed coding agent, reporting large search-efficiency gains while stopping well short of autonomous model-weight self-modification.
OpenAI says it could push math-research capability harder but is prioritising recursive self-improvement and automated alignment research instead, without publishing a formal slowdown trigger.
Most self-improving coding agents rewrite themselves after a single failure, throwing away the archive of everything they have already tried; a new method adds two edit operations that use multiple trajectories and a competing agent's evidence instea
Ouroboros is a coding agent whose tools, prompts and core implementation change through reviewed commits that become the runtime for its next task, and its longest public deployment ran live for 161 days across seven surfaces.
Redwood Research's Ryan Greenblatt told Dwarkesh Patel that once AI matches top human AI researchers the feedback loop could compress four or five years of progress into a single year, and that what models still lack is not deep insight but hands-on
Evo-Bench holds the model and budget fixed and measures only what improving its own scaffolding is worth, finding gains of up to 16.6 points that come close to human-engineered baselines everywhere except tasks with prescribed workflows.
Frontis-MA1, released with full weights and stack, raises its base model's medal average on a machine-learning engineering benchmark from 39.4% to 60.6%, and to 71.2% with a stronger search - all within a 12-hour budget on a single consumer GPU cappe