[doctorpangloss]

> Should Apple be held accountable for that?

You could ask the same thing about gun manufacturers and shootings. This is a totally normative, political question.

And before you start moving the goalposts: an eye opening experience for me was considering that while you can dismiss an individual’s claims of “harms” as imaginary, what about a huge group of people? Specifically, if the authors unionized (they did) and they say that AI training on their work harms them (they do), does that not make in “real,” in a special way, the same special way that a law that comes into being via a popular vote is more “real” compared to a law that is made by fiat by a dictator? I am just trying to open your mind past these really basic sentiments and gotchas.

No matter what, AI developers must grapple with popular opinions about AI.