A survey of generative AI adoption and perceived productivity among scientists who program

Summary

Survey of 868 scientists who program, on GenAI adoption, tool preference and perceived productivity. Adoption is highest among students and less experienced programmers; scientists overwhelmingly prefer general conversational interfaces (ChatGPT) over developer-specific tools; and the strongest predictor of perceived productivity is the number of lines of generated code accepted at once — implying researchers gauge productivity by generation volume rather than validation. A sharp, quantitative deskilling/over-reliance datapoint for the research-pipeline argument.

Contribution

Queued — not yet read.

Method

Queued — not yet read.

Relevance to RISE

Surfaced by the 2026-08-29 literature sweep as a medium-priority candidate for the RISE knowledge base (angle: human–AI research collaboration). Verified against: arXiv abstract page opened 2026-08-29 (title, 4 authors, submitted 22 Dec 2025, full abstract read)

Critique / open questions

Queued — not yet assessed.

Key quotes

Programming is essential to modern scientific research, yet most scientists report inadequate training for the software development their work demands. (abstract)

Generative AI tools capable of code generation may support scientific programmers, but user studies indicate risks of over-reliance, particularly among inexperienced users. (abstract)