Sadly I can't give any specific technical details without potentially doxxing myself to my employer, but I've reviewed 5 or 6 AI-generated PRs in the last year that have contained some pretty severe bugs — the kind of stuff that would have caused serious service outages and security breaches if allowed to go to production.

Those 5 or 6 were among dozens of PRs, and the amount of code I'm being asked to review is only increasing. Even being diligent, I'm sure it's only a matter of time before something nasty slips by me, especially now that the process of reviewing code has become so fucking exhausting.

Yet another reason to not ship code like this. I keep my slop code to myself, tyvm.

I'm so mad that this is "news" to people

As usual HN commenters come running to explain away AI concerns

Papier-maché code

TBF inexperienced developer's generate fragile codebases too.

They say, using an ai generated image

No shit Sherlock. Everyone with a functional brain knew that from the beginning. We also know that the word is license whitewashing by slop, not AI.

Only morons think they can develop good software by using coding agents. And yes, that includes Torvalds, whose sanity is pretty much a goner.

In my experience, most SMEs and quite a few enterprise sized companies had fragile codebases way before AI came along. Many with siloed workers, poor documentation and either weak or overly rigid approval and deployment processes.

It certainly can make code more fragile. It’s a new tool that’s currently being poorly used more often than not.