METR follow-up: many developers withheld tasks; self-reported speedups are unreliable

Oct 26, 2026 · Issue 13 · 4 sources, 5 articles

In METR's 2026 follow-up study, 30–50% of invited developers refused to submit tasks for the no-AI comparison. The articles say self-estimation is a poor measure of speedup.

Who is affected and how

tech lead

Do not base productivity on self-reports, because METR says they are biased.

CTO

Back up the return on AI investment with objective measurements.

Narrative: More code, but not more delivered software

Sources

• 19% slower, and sure they were 20% faster Why I wrote a book about building software with AI agents - DEV Communitydev.to · Oct 26, 2026
• Developers Felt 20% Faster With AI. They Were 19% Slower.towards data science · Oct 30, 2026
• How the Software Engineer Role Is Changing with AI - Ajit Singhsinghajit.com · Oct 30, 2026
• Proton redefines the software engineer's role in the AI erajawlah.co · Nov 1, 2026

Machine-generated summary based on the titles and excerpts of the source articles. The facts come from the linked article.