In addition to the steady stream of LLM-generated pitches we're seeing at LWN, I'm also seeing an uptick in pitches for LLM-generated projects and distributions.
While we do write about projects, like the Linux kernel, that are using LLMs somewhere in the process, I'm not going to be writing about any of these flash-in-the-pan projects anytime soon.
- Generally, I don't write a review of a project that I wouldn't use myself.
- I also am wary of writing about a project until I feel confident it's going to be around for a while, and there's either some community behind it or at least a dedicated maintainer who will likely stick with the project. (Nothing is promised, of course - sometimes things happen.)
- I see no point in spotlighting something that a person slapped together with Claude credits in a few days when other users can do the same thing, and more appropriate to their very specific use case.
A few interesting, perhaps, observations:
- I've been writing about FOSS for 27 years, off and on. Getting actual maintainers to pitch their projects is usually like pulling teeth. I've had so many conversations at events with people who maintain projects or who know about projects to *pitch them* to publications so everybody can learn about them.
- LLMs come along and we get a steady stream of "oh, look at my project".
- Zero of the people who are pitching are up-front about the projects being LLM-assisted or LLM-generated.
I think there's a lesson or three in there somewhere about the impacts and motivations around these tools, what they mean for community, and the downward slope we're already on.
@jzb I maintain a small-to-medium fairly boring open source project (ShinyCMS) with a very low userbase, that has been around decades and been reimplemented in a second language during its existence; where can I read more about whether it might be of interest and how to go about pitching it?
Zero LLM usage! The project originally was started as a learning experience for me, and for the most part that's what it's stayed.