Lenny’s Newsletter · Subscriber post · Building AI products · Metrics, data & experimentation

AI tools are overdelivering: results from our large-scale AI productivity survey

What exactly AI is doing for people, which AI tools have product-market fit, where the biggest opportunities remain, and what it all means

Lenny RachitskyDec 23, 202524 min♥ 451
SourceLenny’s Newsletter
KindSubscriber post
PublishedDec 23, 2025
Readers♥ 451
Originallennysnewsletter.com ↗
N:

Lenny Rachitsky and Noam Segal report on a 1,750-respondent survey of tech workers about AI productivity. The core argument is that AI is overdelivering for most people, with the biggest gains in founders and PMs, while the next opportunities lie in upstream thinking tasks, research, and tedious post-coding work, and agentic adoption remains early.

Subscriber post — summary only

01Key takeaways

  • Most respondents say AI exceeded expectations and saves at least half a day weekly on key tasks.
  • AI helps PMs produce artifacts well but lags on upstream research and roadmap thinking, a clear opportunity.
  • Founders use AI most as a strategic thinking partner, while designers report the smallest benefits.
  • Engineers want AI for documentation, code review, and testing, not just writing code.
  • Adoption of autonomous agents is still low; most workflows remain human-directed despite strong interest.

02Frameworks mentioned

Summary and takeaways written by PM Atlas; quotes are short excerpts. © the original author.