Towards End-to-End Automation of AI Research

Summary

The AI Scientist in Nature (25 March 2026) - the peer-reviewed, high-visibility version of the Sakana AI end-to-end system, covering both a template-scaffolded focused mode and a template-free open-ended agentic-search mode, and reporting a fully machine-generated manuscript that passed the first round of review at an ML workshop with a 70% acceptance rate. Explicitly names the systemic risks of taxing overwhelmed review systems and adding noise to the literature. IMPORTANT: the KB already holds the AI Scientist v1/v2 arXiv system papers, but this Nature article is the canonical citable version and is not in the skip list; add it and cross-link.

Contribution

Queued — not yet read.

Method

Queued — not yet read.

Relevance to RISE

Surfaced by the 2026-08-29 literature sweep as a high-priority candidate for the RISE knowledge base (angle: autonomous AI scientists and their evaluation). Verified against: arXiv abstract page opened 2026-08-29 (8 authors, full abstract read); Crossref publisher record for the Nature version opened 2026-08-29 (Nature 651(8107):914-919, published 2026-03-25, author order Lu, Lu, Lange, Yamada, Hu, Foerster, Ha, Clune)

Critique / open questions

Queued — not yet assessed.

Key quotes

The automation of science is a long-standing ambition in the field of AI. (abstract)

While the community has made significant progress in automating individual components of the scientific process, a system that autonomously navigates the entire research lifecycle -- from conception to publication -- has remained out of reach. (abstract)