In the article they quote someone who said he’s glad his code has been used by LLMs because he’s always been working to democratize tech, but I still struggle to see how this democratized it. A handful of companies control access to the AI and will use that as an excuse to keep more people out of tech. I’m glad he is ok with his work having been used, but plenty of people aren’t and wouldn’t want companies like Microsoft or Google profiling off them without compensation. I wonder if his opinion would change if he was one of the people let go and replaced by AI. If then he could see that this isn’t about democratizing things, but further concentrating wealth in the hands of these large corporations. Additionally, this has reduced friction for bad actors and slop generators. Is he glad his work makes people confident contributing to FOSS projects even if they have no idea what the quality of the code is like increasing overhead time for the people maintaining the project? I think a totally free AI for personal use in a society much better than ours would actually be acceptable, but we don’t live in that society, we live in this one. Any use of AI now is just rewarding these companies for a “move fast and break things” mentality that truly does not care what things end up being broken (the economy, ownership rights, individual families, the environment, professional integrity, trust in the open source community, etc).
I think there are a couple of things here.
I think he’s coping honestly, I’m not sure if he truly believes it’s democratizing anything, he has spoken against AI as well, so it’s weird seeing this contradiction. He has also been very against programming without thinking deeply about the code for so long, again a contradiction. He seems to be coping with all of the work being wasted. All of those years he and others have worked to produce good code, just gone.
The other thing I feel is going on right now is, how corporations kind of took over Linux. At some point they were improving the project, but now they definitely aren’t and he can’t do anything about that. There are many contributions accepted now, despite not being very well done, just because they need to keep going and there’s so much to go through, but there aren’t enough maintainers and probably never will be because of how everything in Linux is organized.
Which brings me to my last point, maintainers. I think he’s hoping that LLMs will be a great tool to help or even replace maintainers, because they can’t deal with the work otherwise.
I think with this whole mess, he’s just trying to keep it together and be pragmatic about it as he’s always done, but this time he’s just losing.
he stopped being a working class person a long time ago therefore he doesn’t give a fuck what happens to programmers. he himself said he hasn’t contributed code in a long time, he’s just a glorified project manager now.
He can do what he wants, he is a free person and has every right to do so.
But i - as an equally free person (albeit with much, much much less wealth) - will never use voluntarily software that is generated by such an abomination and will not stop to advocate against it.
Let me use this opportunity to point towards NetBSD which has a clear “No LLM” clause in its commit guidelines