A cruel irony of coding agents is that everyone who blew off automated testing for the past 20 years is now telling the AI to do TDD all the time.

But because LLMs were trained on decades of their shitty tests, the agents are also terrible at testing.

Justin Searls

AI coding agents face a paradox: developers who historically avoided automated testing now instruct AI to follow test-driven development practices. However, since large language models were trained on decades of poorly written tests, the resulting AI-generated tests are equally problematic, perpetuating the cycle of low-quality test code.

A cruel irony of coding agents is that…

<p>Not my experience, I think it’s awesome. Occasionally I have to delete some overeager tests and rarely I have to ask for something else. But overall I’m very impressed.</p>
