cimbazarov

Luddism part 2.

I think the question is if AI, as a tool, can make such a quantitative change in productivity that there is a qualitative change in the relations of production. Otherwise it's just going to be the same as all other increases in productivity this century which sharpen the contradictions of capitalism.

I mean from my experience with chatgpt, it has def been trained on Marxist theory texts. And it gets some of it right and some of it wrong in the same way it gets anything right or wrong.

LLM's are black boxes. You can't modify their algorithm and influence them at the layer of abstraction to be pro or anti anything (at least from my understanding). The only way you could do that is by cherry picking its training data, and even then it's still a black box so it could potentially have a bias towards the opposite of what you intend.

I've started contemplating if a butlerian jihad is more likely than a proletarian revolution at this point with how little class conscious there is in America.

It is interesting to see that even tech-bros can see some of the contradictions in AI. I do feel it's easier to organize them against AI rather than against billionaires who own it (since in the tech bro mind they are already on the cusp of being one of the billionaires).

Also if I could indulge in some sci-fi speculation for a bit: what if the "AI takeover" follows one of the stories in I, Robot (the book) where there's a official who is an AI but it's not clear to the public that he is, creating this environment of ambiguity. Then we have more "AI's" masquerading as real people (take the recent event with the AI Instagram profiles) until everyone we are surrounded by is an AI before we even know it.

It would be cool to see more tech worker solidarity, but I feel tech workers in particular have the most animosity towards each other, at least in big tech. Of course this is the result of how these companies are structured, but it's hard to get that across to them.