In a move chillingly reminiscent of George Orwell’s 1984, Iowa Governor Kim Reynolds signed Senate File 496 (SF 496), which enacted sweeping changes to the state’s education curriculum and banned books that are not deemed “age appropriate” or that contain “descriptions or visual depictions of a sex act,” per Iowa Code 702.17.
Iowa’s thought police had a problem: how to comply with the law. Their solution: use ChatGPT to identify books that meet Iowa’s book-banning standards. Bridgette Exman, Mason City’s Assistant Superintendent of Curriculum and Instruction, said it was “simply not feasible to read every book and filter for these new requirements,” so the administrators prompted ChatGPT with the specific language of Iowa’s new law: “Does [book] contain a description or depiction of a sex act?”
According to The Gazette, 19 books were pulled from Mason City school libraries including Alice Walker’s The Color Purple, Margaret Atwood’s The Handmaid’s Tale, Toni Morrison’s Beloved, and Buzz Bissinger’s Friday Night Lights.
Leveraging large language models to keep minds small is ironic, and book banning makes my blood boil, but what truly gives me pause is that ChatGPT is not the right tool for this job. The prompt may have been politically (and legislatively) correct, but it was not a prompt that could guarantee results that are aligned with the legislation. So many incorrect assumptions were made here; it is an object lesson in what not to do with ChatGPT.
If you want to go deeper into my technological argument (not the philosophical one) and get a better understanding of when and where using ChatGPT is appropriate, sign up for our free online course Generative AI for Execs. You’ll learn how to use a simple conversational AI decision matrix to ensure the best outcomes when using tools like ChatGPT.
Author’s note: This is not a sponsored post. I am the author of this article and it expresses my own opinions. I am not, nor is my company, receiving compensation for it.