2026-09-29 03:07 UTC
Always worth reading the fine print. e.g., people citing a "study" done by someone at ThoughtWorks that concluded that asking agents to do TDD produced no discernible benefit.
If you read the blog post, they actually failed to get the agent to follow the TDD discipline. Not the same thing.
The real conclusion should be that LLMs don't follow instructions. So stop asking them to.
Replies (1)
-
@jasongorman@mstdn.business 2026-09-29 03:12
I can say with confidence - because there's good data on this (including some of my own) - that specifying with tests dramatically improves results, as does working in micro-cycles with continuous testing, code review and refactoring. The jury's not out on that. But, sure, ask an LLM-based agent to do TDD, and it just won't. *You* have to control the workflow at every step.