AI is contentious among authors. So why are some feeding it their own writing?
LEILA FADEL, HOST:
The literary world has a complicated relationship with artificial intelligence. Many authors worry about the use of their copyrighted material in works generated by AI, but some are using AI to drive creativity. NPR's Chloe Veltman reports.
CHLOE VELTMAN, BYLINE: The vast majority of authors don't currently use artificial intelligence as part of their creative process, or at least don't admit to it. But according to a recent poll from the writers' advocacy nonprofit the Authors Guild, 13% say they do use AI for things like brainstorming character ideas and creating outlines. Chris Anderson says he wanted to push the technology further.
CHRIS ANDERSON: See whether in fact it can do more than just help me organize my thoughts, but actually start injecting new thoughts.
VELTMAN: Anderson is best known as the author of technology and business-oriented nonfiction books like "The Long Tale." Lately he's been trying his hand at fiction and is at work on his second novel about drone warfare. Anderson says he fed parts of his first novel into an AI writing platform to help him write this new one. The system surprised him by moving his opening scene from a corporate meeting room to a karaoke bar.
ANDERSON: And I was like, you know, that could work. I ended up writing the scene myself, but the idea was the AI's.
VELTMAN: But Anderson says he didn't use a single word the AI platform generated. The sentences were grammatically correct but fell way short in terms of replicating his writing style. And the author says he's OK with that.
ANDERSON: Maybe that's just the universe telling me that the act of writing actually involves writing.
VELTMAN: It's currently very hard for off-the-shelf AI models like GPT and Claude to emulate contemporary literary authors' styles. The authors NPR talked with say that's because these models are predominantly trained on content scraped from the internet, like news articles and how-to manuals - standard, non-literary prose. But some authors, like Sasha Stiles, say they've been able to make these systems suit their stylistic needs.
SASHA STILES: There are moments where I do ask my machine collaborator to write something and then I use what's come out verbatim.
VELTMAN: The poet and AI researcher says she wanted to make the off-the-shelf AI models she'd been experimenting with for years more responsive to her own poetic voice. So she started customizing them by inputting her finished poems, drafts and research notes.
STILES: All with the intention to sort of mentor a bespoke poetic alter ego.
VELTMAN: Stiles says she has collaborated with this bespoke poetic alter ego on a variety of projects, among them a multi-media poem last year for luxury fashion brand Gucci.
(SOUNDBITE OF ARCHIVED RECORDING)
STILES: How many years have come and gone? How many loves have come and gone?
VELTMAN: Stiles says working with her AI self has led her to ask questions about whether what she's doing is, in fact, poetic and where the line falls between the human and the machine.
STILES: It's been really a provocative thing to be able to use these tools to create poetry.
VELTMAN: These types of experiments are also provocative in another way. Authors Guild CEO Mary Rasenberger says she's not opposed to authors training AI models on their own writing.
MARY RASENBERGER: If you're using AI to create derivative works of your own work, that is completely acceptable.
VELTMAN: But building an AI system that responds fluently to user prompts requires vast amounts of training data, so the foundational AI models that underpin most of these investigations in literary style may contain copyrighted works. Rasenberger points to the recent wave of lawsuits brought by authors alleging AI companies trained their models on unauthorized copies of articles and books.
RASENBERGER: That creates some real ethical concerns if the output does in fact contain other people's works, because that you should be getting permission for.
VELTMAN: Award-winning speculative fiction writer Ken Liu says he wanted to circumvent these ethical problems while at the same time create new aesthetic possibilities using AI. So the former software engineer and lawyer attempted to train an AI model solely on his own output and nothing else.
KEN LIU: All my short stories, all my novels.
VELTMAN: Liu says he knew this approach was doomed to fail. That's because the entire life's work of any single writer simply doesn't contain enough words to produce a viable so-called large language model.
LIU: I don't care how prolific you are, it's just not going to work.
VELTMAN: Liu's AI system built only on his own writing produced predictable results.
LIU: It barely generated any phrases even. A lot of it was just gibberish.
VELTMAN: Yet, for Liu, that was the point. He put this gibberish to work in a short story. "50 Things Every AI Working With Humans Should Know" is a meditation on what it means to be human from the perspective of a machine. Here's a robo voice reading a short clip.
AUTOMATED VOICE: Man reached the torch for something darker perified. It seemed the billboding (ph). Not full of pain facio in bed.
VELTMAN: And Liu continues to experiment with AI. He says the technology is still very limited, but it has promise.
LIU: So what is the point of experiment with AI? The point for me really is about pushing the boundaries of what is art.
VELTMAN: He says, so far, these endeavors have reaffirmed why human art matters, that ultimately AI is just another useful creative tool like a paintbrush or a camera.
Chloe Veltman, NPR News.
Copyright © 2024 NPR. All rights reserved. Visit our website terms of use and permissions pages at www.npr.org for further information.
NPR transcripts are created on a rush deadline by an NPR contractor. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.