Treat Agent Output Like Compiler Output

Should be tagged with ai as well.

Arguments don't hold up, it's the same argument patterns, and it's tiring.

AI-checks-AI pipelines as first-class CI infrastructure, not bolt-on curiosity

The solution for 'too much AI' is 'more AI' coming from a company that sells AI.

Here's the contrarian edge: the engineers most resistant to lights-out codebases are often the same engineers who would resist, if they could time-travel back, the idea that you don't need to review your linker's output. The intuition feels protective. It is actually just unfamiliarity with where the trust has been relocated.

Blame the developers and engineers for 'being stupid' and 'just being scared'. Let's not address their concerns, just call it stupid and move on.

Hardware chip companies already work with black-box components verified by acceptance tests rather than human review.

Shoving AI in the same category as fully deterministic processes that have been verified is just...I'm really trying to be nice here.

The compiler didn't make us stop caring about correctness.

Yes it did, look at how many projects were forced to have AI policies due to poor submissions and the overwhelming low quality code and reports. If the author wants to look the other way, they are more than welcome to, but that statement is just disrespectful to all the devs out there trying to cope.

The scary part isn't the lights-out codebase. The scary part is how few teams are treating what replaces the review as serious engineering work.

It is. It is not.

I usually don't see posts like this on Lobsters, but wow. It's undermining an entire group of professionals' lived experience. This is sad.