Having already assured readers that a significant body of copyrighted books were used without permission to train large language models as part of the corpus researchers call Books3, the Atlantic is now ready to collect your rage clicks with a searchable database of those titles. “Since my article appeared, I’ve heard from several authors wanting to know if their work is in Books3. In almost all cases, the answer has been yes.” They tabulated 206 works by Nora Roberts; 133 from Danielle Steel; and only 121 by James Patterson. Amusingly, the publication also wants us to know that a long […]
