When bestselling thriller author Douglas Preston started playing with OpenAI’s new chatbot, ChatGPT, he was initially impressed. But then he realized how much in-depth knowledge GPT had of the books he had written. When prompted, he provided detailed plot summaries and descriptions of even minor characters. He was convinced that he could only achieve this if he had read his books.
Large language models, the type of artificial intelligence that underpins programs like ChatGPT, don’t come into the world fully formed. They must first be trained on incredibly large amounts of text. Douglas Preston and 16 other authors, including George RR Martin, Jodi Piccoult and Jonathan Franzen, were convinced that their novels had been used to train the GPT without their permission. So, in September, they sued OpenAI for copyright infringement.
This sort of thing seems to be happening frequently lately: one giant tech company or another “move fast and break things,” exploring the boundaries of what might or might not be allowed without first asking permission. On today’s show, we try to make sense of what OpenAI would have done by training its AI on massive amounts of copyrighted material. It was good ? Was it bad? Was it legal?
This episode was hosted by Keith Romer and Erika Beras and produced by Willa Rubin and Sam Yellowhorse Kesler. It was edited by Kenny Malone and fact-checked by Sierra Juarez. Engineering by Robert Rodriguez. Alex Goldmark is the executive producer of Planet Money.
Help support Planet Money and get bonus episodes by subscribing to Planet Money+ in Apple Podcasts or to plus.npr.org/planetmoney.
Always free on these links: Apple Podcasts, Spotify, Google Podcasts, NPR One or wherever you get podcasts.
Find more from Planet Money: Facebook / Instagram / Tic Tac / Our weekly Newsletter.
Music: Elias Music – “Elevated”, Universal Music Production – “Don’t Cross the Line” and “This is Not Goodbye”