Artificial intelligence may be the future. But is it the future of books?
In recent weeks, AI-generated content from OpenAI’s model ChatGPT has been sweeping the internet on sites like Twitter as users gawk over the hilarious absurdity—and the often shocking prescience—of the automated output. Trained on related texts or images and given a prompt, AI models like ChatGPT and DALL-E can generate entire pages of text or imagery that oftentimes seem distinctly human-made.
We wondered, what would an AI-generated book proposal look like? Our social media manager sent a prompt to the OpenAI ChatGPT chatbot: “Please write me a pitch letter for the next bestselling MIT Press book.”
The resulting pitch was cogent (and polite, we must note, beginning with a kindly “Dear [Editor],”)—if a little clunky and rudimentary at times. The topic of this bestselling book, ChatGPT writes, is “the many ways in which technology is shaping our worlds and our lives.”
“In addition to discussing the latest advancements in technology, I will also delve into the potential drawbacks and ethical concerns that come with these developments,” the pitch continues. “As the world becomes more reliant on technology, it is crucial that we consider the impact it has on our society and our individual lives.”
We considered several AI-generated cover options for the book proposal as well. Dark and sinister, they seemed more appropriate for a science fiction dystopian title than for the nonfiction work in question. Nevertheless, the AI’s proposed table of contents presents a sweeping exploration of the very issues many scholars and policymakers grapple with today.
So, should we sign this book?
Phil Laughlin, senior acquisitions editor at the MIT Press, critiqued the pitch: “Too wordy for me. I like my pitch ‘letters’ to be three sentences: ‘I’m Professor So-and-So at Such-and Such-University. I’ve written a draft manuscript on topic X. Would you like to see it?’ Google will answer any questions I might have.”
This particular example notwithstanding, AI programs have been tasked with generating creative work with mixed results. “We need to set aside the old stereotype that computer programs simply follow fixed rules and do what humans have programmed them to do, and so lack any capacity for creativity,” argues Keith Holyoak, author of The Spider’s Thread, in an MIT Press Reader article adapted from his book.
Holyoak considers AI-generated poetry to be an area where computers don’t meet human standards. Consciousness—which AI bots still lack—seems integral to creating something artistic and original, such as a book, cover, or poem. The key difference between a human poet and an AI poet, Holyoak writes, is this inner experience.
“The absence of inner experience means that AI lacks what is most needed to appreciate poetry: a sense of poetic truth, which is grounded not in objective reality but rather in subjective experience,” Holyoak argues. “What AI has already accomplished is spectacular, and its further advances will continue to change the world. But for all the functions an AI can potentially achieve—the ability to converse intelligently with humans in their natural languages, to interpret their emotions based on facial expression and tone of voice, even to create new artistic works that give humans pleasure—an intelligent program will fall short of authenticity as a poet.”
Back to our book pitch; it seems that we won’t be signing an AI-generated book anytime soon. But we certainly won’t say never.