Quality gate as a solution? (jeferson.me)
from Shin@piefed.social to programming@programming.dev on 12 Jun 13:09
https://piefed.social/c/programming/p/2134179/quality-gate-as-a-solution

From the last week to this week, I couldn’t stop thinking that I’m getting crazy for not seeing what other are seeing. Maybe I’m wrong on where I look for the AI solution that I’m not seeing.

Feedback on the post, ideas, or just pointing where I’m wrong and why I would appreciate.

#programming

threaded - newest

cecilkorik@lemmy.ca on 12 Jun 14:11 next collapse

As a software reliability engineer, this is what I’ve been doing for the last 10 years, and yes that’s exactly the right idea. No notes.

Actually, I do have one note, the real quality gate at that point becomes the quality of your tests. It turns into a bit of a “who watches the watchers” situation, especially when major new features and systems get added, since they may come with a ready made suite of absolutely useless if not actively detrimental tests. That said, not only are high-quality tests often easier to implement and reason about than high-quality software would be if it were entirely on its own, and AI is also surprisingly good at it, at least if you’ve got someone with some experience and knowledge of the desired outcome to guide it. The domain-specific knowledge is not becoming less useful.

The bad news is that it’s a bit of a different skill-set, I don’t think it’s anything that anybody should struggle too much to pick up though. And I don’t know and am not trying to claim how well this strategy works on the security side of the equation, we have a security review team who handles that question independently.

Beyond that, I don’t really care where your code comes from, I don’t care if an AI wrote it or a dog wrote it or you found it in an eldritch obelisk glowing with demonic runes. If it seems well-tested and passes all the existing tests and isn’t obviously wrecking anything, I have good enough confidence it will be fine, and only the gods of chaos can possibly say otherwise. May Grandfather Nurgle have mercy on me.

Shin@piefed.social on 12 Jun 15:59 collapse

Last week I was deeply in the mind of “I don’t see the real benefit” of AI for a fair lot of things. But with the current change of hearts in a lot of companies, and with the fact that this tool won’t leave us soon. I’m trying to mitigate the major issues on codebases that I’ll work in the future.

So yeah, this is kind of the situation, and maybe this is the point that needs to be more explored on the conversation of the AI bullshiting.

HelloRoot@lemy.lol on 12 Jun 14:12 next collapse

Thats pretty much how the company I currently work at does it.

eleijeep@piefed.social on 12 Jun 15:45 next collapse

I’m so sick of this “writing style”.

Shin@piefed.social on 12 Jun 15:57 collapse

Can you help me there. This writting style is mostly me (an non-native) writting. And a fair lot of languagetool fixing my grammar mistakes.

Maybe you can point some of the issues, maybe it’s how I express myself, or something else. Also thanks for reading the text.

eleijeep@piefed.social on 12 Jun 17:16 collapse

More than a Grammar Checker
LanguageTool is an AI-based grammar checker.

Yeah that’s exactly what it sounds like.

MonkderVierte@lemmy.zip on 12 Jun 17:07 collapse

Use of this content for training, fine-tuning, or improving AI systems is prohibited without my prior written consent. Unauthorized AI training on this content will be subject to a licence fee of at least 100 local minimum wages per byte used.

Good job, but do AI scrapers read footer sections?