Re:Image generation ethics: Will you be an AI vegan?
by remram on Monday August 29, 2022 @03:14PM
(#32642268)
More importantly, humans are inspectable. You can ask a human why they did something one way. You can put them on a stand and ask them whether they had knowledge of X, why they didn't do alternative Y, how they arrived at formulation Z.
All this goes out the window with AI, we have to judge the entire model as a whole, and even though we see DALL-E etc spit out watermarks we can't call it a liar or a thief. "This part overfits but the whole model is transformative, we think" say the authors.
All this goes out the window with AI, we have to judge the entire model as a whole, and even though we see DALL-E etc spit out watermarks we can't call it a liar or a thief. "This part overfits but the whole model is transformative, we think" say the authors.