I don’t really understand using AI for code reviews. Shouldn’t it be the other way around?

I have my setup hooked up to SonarQube and compell it to run all its changes through that. I prefer a static analysis tool with predefined checks to keep the agent in line. I just have it do a TDD workflow using that as part of the refactor step.

It’s made my work life a lot easier, but do I need it? No. At the end of the day the only thing that would change is that things take longer to get done.

Even with AI, I still read the docs to understand what it wants to do. Half the time I need to challenge it or refine the criteria.

The one place it has been a godsend for me is working in unfamiliar areas (a new library, new code ase, etc). Helps me onboard so much faster. Has helped accelerate learning too. But do I NEED it. No I don’t. I’m still smarter than it, and all that’s needed is time.

I have mixed feelings about that. Thanks for explaining!

So why does everyone hate lemmygrad and hexbear? I’m out of the loop on that.