Working Surfacelink · 15 Nov 2026
02
A steam traction engine pulling a plough across an open field, its driver seen from behind at the wheel with a repair toolbox strapped beside the seat, one long furrow trailing behind, the whole scene drawn with wide empty margins on every side.
InfoQ

Doing The Right Thing: Accountability In The Age Of AI Coding Summary

Shane Hastie with Hugo Venturini · 13 November 2026

Company story · critique & review · roles · Human approval

Read the original
Takeaway

Hugo Venturini, a former Meta engineering manager, told the news site InfoQ that whoever directs an AI tool answers for its code, and teams may choose which changes need review.

Summary

Shane Hastie, lead editor for culture and methods at InfoQ, a software engineering news site, interviewed Hugo Venturini on the InfoQ Engineering Culture Podcast. Venturini, a former manager at Meta and former chief technology officer of the search engine Qwant, now works at a startup that makes tooling for AI. The conversation covers who answers for code written by AI, how much of it needs review, and how leaders build a culture where people accept that responsibility.

Key points
  • Venturini said that people remain accountable for AI-written code, and that the person directing the tool should answer when it fails.
  • Large pull requests predate AI, and the team Venturini works in does not review every change. An engineer who merges without review answers for fixing it.
  • Reviews are kept for the core of the product, both for quality and to share knowledge across the team.
  • Venturini linked Microsoft's 2014 merger of the engineer and tester roles to higher quality, and said that blaming the AI tool risks reversing that gain.
Implication

An engineering team can write down which kinds of AI-written change an engineer may merge alone, and who fixes them if they break. This is an inference and is not Venturini's claim. The piece is one leader's view from an interview, with no team data.

Plays
Engineering

Let engineers merge routine changes without a second reviewer when they accept responsibility for fixing them, and keep reviews for the core of the product.

Derived by Working Surface from the article; more in the Playbook. Source line: we don't see an increase of quality yet, and I believe it's because people are still blaming the AI and don't feel accountable for what the tooling is doing.

Source issue

15 November 2026: A team can rely on an AI agent's work only when the check sits outside the agent and someone other than the builder decides what correct means.