Slate: “Over the coming years, A.I. companies will release even more advanced models that will remind us that this is just the beginning. At least one of these tools will be different in an important way: It will be prohibited from seeing 80 million of the images that helped teach its predecessors to draw and paint. If you think of A.I. image training sets as lesson plans and word-to-image tools like DALL-E and Stable Diffusion as college students, it’s kind of like saying that incoming freshmen are prohibited from taking one of the outgoing seniors’ core requirement classes. Why? Because two Berlin-based musicians persuaded the head of a multibillion-dollar company to give artists more power. Many illustrators, designers, and photographers are furious that their work had been scraped from the internet to train an A.I. Last September, Mathew Dryhust, an artist and academic, and the experimental electronic artist Holly Herndon, created HaveIBeenTrained, a tool that enables you to see if your images have been used and then request that future A.I. models avoid them. It recently became clear that this was more than a subversive art project when the CEO of the company behind Stable Diffusion, one of the most widely used A.I. image generators, gave people a March 3 deadline to opt out using Herndon and Dryhurst’s tool. Millions of requests poured in, Dryhurst told me. As of this week, around 40,000 artists and many companies, including Shutterstock, a major image distributor, have opted out.