Thoughtworks essay: experienced developers wary of agentic coding
Original title: Why are good developers so wary of agentic development?
A no-human-code week still left a frequently red build and most human time at the whiteboard.
Rachel on Martin Fowler’s site argues that experienced developers hesitate to let coding agents do all execution because production confidence, boundaries, and failure handling are unproven. A Thoughtworks week in which agents wrote all code for an airline-style application reached about 80 percent completion, with humans spending roughly two-thirds of their time at the whiteboard and one-third on the harness. The application ran, but the build was frequently red, which the author says undermines the signal continuous integration is meant to provide.
Unconfirmed: The two-thirds and one-third time split is the team’s own estimate.
- Humans wrote no code; agents produced it, including outside working hours.
- The team estimated about two-thirds of time at the whiteboard.
- A product person committed more feature code than other team members.
- The build stayed red often enough that failures could be ignored.
- Organizations
- Thoughtworks
From the original article
9 min read at martinfowler.com
Good developers’ caution about agentic development comes from knowing that built isn’t the same as working in production. I’ve been surprised by how many excellent developers have been hesitant to go all in on coding agents. These are people who usually try new ways of working before anyone else and in many cases invent new tools and practices that support building the application well.
By torutofu · martinfowler.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.