, which allows artists to see if their work was used to train popular A.I. art models.
. There are major lawsuits, not just by her, but by celebrities like Sarah Silverman. The administration just isn’t listening.”against OpenAI and Meta over claims the companies violated her copyright by training their chatbots ChatGPT and LLaMA on the text of her book“We represent American creators—including authors, visual artists, and programmers—whose work has been misused by A.I.
Ben Zhao, a professor of computer science at the University of Chicago, is one of the professors who led a research team that developed—a technology that allows artists to prevent A.I. platforms from stealing their artistic style by making subtle shifts to brushstrokes and palette. The “voluntary” nature of these commitments renders these commitments “meaningless,” he added, calling the provisions outlined in the deal “poorly defined goals” which involve technical problems that lack solutions or may be completely insolvable.
“The assumption that big tech will do the ‘right’ thing despite the obvious financial disincentives is naïve,” he said. The European Union’s laws on text and data mining and a prospective law on A.I. are “much clearer” about providing guidance on issues like data harvesting and model transparency, Dryhurst said. He called the ability to opt out of training models the “most plausible demand” artists can make.