So Ai companies aren't responsible for the training data they stole, aren't responsible for the output. What exactly are they responsible for other than the (limited) profits? Seems to me the only thing they care about is protecting power and the status quo.

What limited productive use they have seems to be constrained to software development. Any kind of deep insight is classified as a hallucination and removed. It's becoming clearer and clearer that these products are a stopgap only; billions of dollars to make sure nothing changes. Yet it will happily spit out obviously fake sources, a different definition of "hallucinations," in the domains of science and law. These are information weapons against small businesses and individual autonomy.

I don't think it's any accident that Character.ai is targeted at children.