Widelands is a free and open source Settlers-like strategy game, and their developers appear to be setting a firm stance against an AI generated contributions.
No, I did not have to do that. Why do you ask?
Just saying that the sentiment I’ve seen expressed by open source developers towards gen AI ranges from cautious optimism to banning it. I’ve yet to see any open source team express that gen AI has been useful to them.
Realistically, it’s because they don’t like wasting their limited (usually unpaid) time reviewing untested AI PRs which may or may not do something useful for the project, publicity is irrelevant.
Attention is currency. I’m learning about the existence of a lot of projects simply because they made anti-AI announcements
Again, I’m not assuming the worst of these devs or any other. I’m just saying you’re more likely to hear about it because everyone knows saying “AI good” will only garner hate for your project and “AI bad” will get you an article or 3
Well, you don’t really need to announce anything, if the AI-generated submissions were super helpful anyways.
But yeah, I guess, all I can say is that I really don’t believe your theory. Especially Widelands could’ve done so many other things in the past, if they cared so much for attention.
But I have also been in the maintainer role, having to deal with generated submissions, and it really isn’t fun. I’m talking specifically about fun, because these are community-driven projects, so you need volunteers to have fun for anything to happen.
In theory, a generated code submission could bring useful changes to the project, but it still isn’t fun to review, because there isn’t a human on the other side that you can teach. Even worse, you’re effectively just talking to an LLM through a middleman. If I wanted to use an LLM, I’d use it directly.
I think you’re taking what I said too negatively toward the developers themselves I put the virtue signalling line in there to hopefully make that clear I wasn’t accusing them of anything. I was more commenting on what kind of news gets picked up by content farms.
I’ve learned about sooo many opensource projects because the devs made some anti-AI comment that got reported here
Each time the team behind an open source project talks about AI, it’s carefully optimistic at best. But more frequent are news of AI being banned.
Probably because open source and enshittification don’t belong together
Have you had to sort clearly AI genned PRs? Just curious
No, I did not have to do that. Why do you ask? Just saying that the sentiment I’ve seen expressed by open source developers towards gen AI ranges from cautious optimism to banning it. I’ve yet to see any open source team express that gen AI has been useful to them.
It’s more frequent because making a “NO AI” announcement is cheap publicity.
I won’t go as far as to say it’s virtue signalling but…devs aren’t dumb, they know these game news content farms are going to pick up the story
Realistically, it’s because they don’t like wasting their limited (usually unpaid) time reviewing untested AI PRs which may or may not do something useful for the project, publicity is irrelevant.
Community-driven open-source projects don’t have anything to sell. They don’t care terribly much for the publicity.
Attention is currency. I’m learning about the existence of a lot of projects simply because they made anti-AI announcements
Again, I’m not assuming the worst of these devs or any other. I’m just saying you’re more likely to hear about it because everyone knows saying “AI good” will only garner hate for your project and “AI bad” will get you an article or 3
Well, you don’t really need to announce anything, if the AI-generated submissions were super helpful anyways.
But yeah, I guess, all I can say is that I really don’t believe your theory. Especially Widelands could’ve done so many other things in the past, if they cared so much for attention.
But I have also been in the maintainer role, having to deal with generated submissions, and it really isn’t fun. I’m talking specifically about fun, because these are community-driven projects, so you need volunteers to have fun for anything to happen.
In theory, a generated code submission could bring useful changes to the project, but it still isn’t fun to review, because there isn’t a human on the other side that you can teach. Even worse, you’re effectively just talking to an LLM through a middleman. If I wanted to use an LLM, I’d use it directly.
I think you’re taking what I said too negatively toward the developers themselves I put the virtue signalling line in there to hopefully make that clear I wasn’t accusing them of anything. I was more commenting on what kind of news gets picked up by content farms.
I’ve learned about sooo many opensource projects because the devs made some anti-AI comment that got reported here