Have you maybe possibly previously been exposed to the concept of an argumentation straw man? Feeding actual works of art into an approximation machine, and no expecting the output of said machine to not be owned by the author of the art is making a big assumption I think. There is the word copy in "copyright" and the model did definitely got a copy of the original at source. No matter the dilution, copyright is being breached, as I understand it.

Thats a reducto ad absrudium at best. While you have a point, schools and even museums are generally compensated for providing these training models to the public, to look at it in a ml way.

Sure, I also had to pay for the books I studied from, but Dr. Tanenbaum is yet to knock at my door to assert copyright on all the code I have written.

This is a perfect example because, depending on the apples you're using, growing them may have required a license and adherence to licensing requirements.

https://mnhardy.umn.edu/apples/licensing https://provarmanagement.com/cosmic-crisp/

Out of curiosity, how would you feel if someone fed your HN comment history into a ML model, then used that to respond on every HN topic and conversation under the username "othergpderetta"?

My HN history is public, so I wouldn't have a problem with training a model with it. I would have a problem with the model attempting to pass as my self of course.

Interesting argument, but it'd be outputting fully sounding word salad garbage.

So exactly like my comments!!

>Excuse me when I make an apple pie from scratch.

You’ll have to invent the universe first.

I expect a big bang anytime now!