Artificial intelligence tools expand scientists’ impact but contract science’s focus

Summary

The flagship peer-reviewed evidence on what AI does to science as a system: 41.3 million natural-science papers, ~311k AI-augmented. Individually, AI-using scientists publish 3.02x more, are cited 4.84x more and lead projects 1.37 years earlier; collectively, the number of topics studied shrinks 4.63% and engagement between scientists falls 22%, producing 'lonely crowds' concentrated in data-rich areas. The single strongest citation for the homogenisation/narrowing-of-science argument and an essential counterweight to productivity-only framings.

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: research productivity of AI). Verified against: Semantic Scholar publisher record opened 2026-08-29 (Nature 649:1237-1243, pub. 2026-01-14, 4 authors) plus the U. Chicago DSI research summary opened for the reported effect sizes; Nature full text is behind an IdP redirect

Critique / open questions

Queued — not yet assessed.

Key quotes

—