A new study from researchers at MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) suggests that the more data a generative model is trained on, the less likely an AI-generated image is to be traced back to a source image. MIT professor David Gifford believes the results bear on the legal question of whether models' outputs are merely derivative works, with implications for copyright. "One way to think about this is that these models are creative," he said a statement.. "They are not simply copying what they are fed, but creating brand new outputs."

  While the illustrator Brad Holland may not be a household name, you could say that if anyone represented the power of illustration to provoke, over the past half century, it would be this artist. Entirely self-taught, he worked his way up from studio rat, sweeping floors in a Chicago tattoo parlor to being one of the first in The New York Times’ roster of artists for its new Op Ed page, in 1970. The Watergate scandal offered the kind of material that excited Holland’s concept...