And my position on AI is that - on top of everything else - it is fundamentally antithetical to, and corrosive to, the most important thing that free software and open, shared knowledge can create and embody and empower, that possibility of durable community and a meaningful common good. It's a tool for enclosing, hobbling and privatizing what community means, and what it could be.
@mhoye A thing I've been trying to figure out for some years is how to get people to think of an AI model as a form of lossy compression. Potentially *very* lossy.
AI datasets are valuable; AI models are datasets in a form you can't reconstruct them without very bad errors. And the more people run AI models, the harder it will be to assemble datasets that don't have errors via nondisclosed AI outputs shuffled in.
@mcc I've had pretty minimal success with "ever had someone send you a picture that's just a screenshot of a screenshot of a screenshot, and it's so blurry and glitchy that it's hard to figure out what you're seeing? All these chatbots are doing that to knowledge and understanding."
It's a difficult thing to convince people to even recognize.
@jplebreton @mhoye my understanding is phone cameras are now adding this technology :(
(and I don't think it's an accident that the hobbling and privatizing of a community of practice is happening before that community can recognize it and organize to protect itself. If it happened afterwards, we already have a term for that: strikebreaking.)