If corporations are a form of slow AI, then what @pluralistic calls #enshittification is indicative of an "alignment problem", when the objective function (paperclip maximization) is persued by destructive means. And like all alignment problems, it has to do with a lack of, or the wrong, constraints.
Add comment