RE: mastodon.gamedev.place/@eniko/Eniko is correct and eloquent and sweary in the thread.It writes code approximately as good as Stack Overflow examples.Which scales about as well as that would.You only run into trouble when you want more than a few lines of code written to illustrate a point; and instead want it to be effective, scalable, secure, legible, maintainable and efficient.It is reviewing the pull requests at work, and that is definitely catching bugs. Some of them even important ones. It is also definitely wasting time on chasing non-bugs.It is possible the ridiculously detailed error coverage in the code is even better code, in some sense. If it matters how you crash and what error you report when you do so and you don't mind reams and reams of error-handling.It may improve. I think I do have a responsibility to be aware of what it can and can't do. If only to fight the people who seem to want to pipe it into root on the prod servers, just for upgrades.And to know when or if it becomes good.I've worked with worse coders, but they also improved. We'll see how well Claude can generalize from textbook examples to effective, scalable, secure, legible, maintainable and efficient over time I guess.I wouldn't rule it out, but there isn't nearly as much training data.#ai #programming
@pre I’m trying to find a workable balance between generate all the code and have no idea how it works, and do have an idea how it works. Yes, there’s a lot of boiler plate that we can gen. But I also notice that the gen slope is covered in soap, and the more I gen, the less I read, all the way beyond the boiler plate, all the way to someone asking “why did you implement it like that” and I have to answer “let me ask my agent” which, I hope you agree, is unworkable?