LLMs use in this repository

[jman-schief]

ref. #5129 (comment)

After observing a sudden spike of comments in issues I have subscribed and seeing massive commits authored by Anthropic's LLM, I start feeling unsure this project.

@Alexays could you please share a bit about how LLMs are being used w.r.t.:

1. the codebase
2. interaction with users (issues and PRs)

I understand the problem of managing the volume of issues but I admit I am not happy that an LLM is proxying the maintainer of this repository, a project I am happy to use but - and I mean no disrespect - lately seems to be tainted by large language models and this makes me unsure and at unease. I hope you can understand my position.

Thank you if you can share a bit more about this topic, I think it's also in the interest of other contributors.

[Alexays]

Since the goal was by no means to maintain the project solely using LLM, the aim was to catch up on the backlog of issues and improve the quality of testing for the project.

[jman-schief]

Well, as I mentioned, massive commits co-authored-by an LLM are not exactly inspiring trust about being properly reviewed. Letting an LLM answer to user issues is equally unsettling. We see that happening to many other FOSS projects, a decline in quality (or trust) due to the influx of LLM generated code and that's a real concern. I am not saying this is happening here, but I would feel more comfortable to know that a human is always firmly at the steering wheel and that is scrupulously reviewing patches, thus my two questions. Just in case, I've pinned the current Debian version of Waybar and wait before upgrading in the future.

Also, I would like to keep this conversation open, maybe other users have opinions to share.

Thanks