HarperCollins and AI: A Mixed Bag of Contracts and Contradictions
Well, well, well, folks! Gather ’round for the latest news from the unpredictable world of publishing! HarperCollins, one of the heavyweight champs of the book biz, has decided to partner up with a mysterious tech company (because who doesn’t love a good cliffhanger?) to sort through some of its nonfiction backlist titles. And what’s the deal? They’re gonna let this tech company use these books to train their shiny new generative AI models. That’s right! We’re officially living in a world where books are being whipped into shape for a big ol’ tech upgrade. Who needs to pick up a dusty old tome when a machine can churn out pithy quotes at the click of a button, eh?
The tech company—unnamed, of course, because anonymity can be quite the fashionable topic—offered a hefty sum of $2,500 per selected book for sponsorship that’ll last up to three years. I don’t know about you, but I suddenly feel like I’ve stumbled upon a bizarre yard sale where classics are being sold like electronic scrap! “$2,500 for Hemingway? What a steal!” It’s like Jurassic Park, but instead of dinosaurs, it’s your average paperback getting a makeover into a digital dino.
The Great AI Debate
Now, HarperCollins claims this partnership is about “model quality and performance,” which is code for, “Let’s make sure our chatbots can finally pass that Turing Test without turning into a glorified auto-reply.” But not everyone in the industry is on board. Some writers are raising their eyebrows (and their stakes). Take author Daniel Kibblesmith, for example. He tweeted, “I’d probably do it for a billion dollars.” Well, if we’re throwing numbers around, I’d like to offer my own book for a million bucks, but I wouldn’t exactly hold my breath on that one! Kibblesmith’s point is hard to dispute—if generative AI is the future, shouldn’t *he* be riding the wave instead of standing on the sidelines with a forlorn look?
But let’s not mince words. HarperCollins isn’t the first to dance with the devil—uh, I mean tech companies. This trend is becoming something of a waltz through the publishing world, with other publishers hitching rides on similar deals. There must be something in the tech water for publishers to pursue this AI tango.
Who’s in the Driver’s Seat?
Giada Pistilli, the head of ethics at Hugging Face (an admirable title if I ever heard one), believes while these accords are a step forward—hey, at least they’re finally offering some cash to publishers—the authors themselves are left clutching at straws with little room for negotiation. That’s like winning half the lottery but being told you can’t spend it. Quite the ethical puzzle! Julien Chouraqui from the French publishing union also weighed in on the situation, expressing that the deals represent “progress.” Progress, indeed! But let’s not kid ourselves—it’s the publishing world’s equivalent of half-arsed effort.
And here’s the kicker small businesses—erm, tech companies really might have no choice but to open their wallets. As they scrape the bottom of the barrel for quality content, their options are thinning out like Luftwaffe rations in a WW2 film. Chouraqui points out that the more they delve into the wild web, the more they’re likely to stumble upon a mix of pirated and legally questionable materials—an every-which-way box of chocolates, if you will. And trust me, you do NOT want to end up with the coconut filling.
Conclusion: A Story Still Unfolding
So, where does that leave us? In a humorous yet precarious balance of technology and literature, where the lines are blurred, and authors are being asked to share their work for a handful of change while the tech giants giggle all the way to the chatbot bank. In this brave new world, who’s truly in charge? Are we feeding our literature to AI harmfully, or is it paving the way to a new frontier? One thing’s for sure—this is a conversation that is far from over, and I can’t wait to see where the next chapter leads.
HarperCollins, one of the largest publishing houses in the United States, has entered into a noteworthy agreement with a currently unnamed technology firm, granting permission to utilize a selection of its books for the purpose of training generative artificial intelligence models.
The tech company outlined in a letter acquired by AFP that it is proposing a monetary compensation of $2,500 for each book chosen to enhance their large language model (LLM), with the contract lasting for a period of up to three years.
AI models necessitate extensive datasets composed of text to refine their capability in understanding and generating everyday language.
“HarperCollins has reached an agreement with an artificial intelligence technology company to allow limited use of select nonfiction backlist titles for training AI models to improve model quality and performance,” the publisher elucidated in a formal statement.
The company emphasized that the agreement has a “limited scope and clear guardrails around model output that respects authors’ rights,” ensuring creators’ interests are taken into account.
The publishing community has reacted to this offer with mixed feelings, as evidenced by authors like Daniel Kibblesmith, who responded to the offer with clear discontent.
“I’d probably do it for a billion dollars. I’d do it for a sum of money that wouldn’t require me to work anymore, since that’s the ultimate goal of this technology,” Kibblesmith commented on the Bluesky social network, showcasing his skepticism regarding the arrangement.
While HarperCollins is among the most significant publishers to finalize such an agreement, it is noteworthy that it is not the first publisher to engage in this type of collaboration.
‘A broader conversation’
Giada Pistilli, the head of ethics at Hugging Face, a prominent French-American open-access AI platform, opined that these agreements signify a positive advancement since they incorporate payments to publishers. However, she expressed concern that the arrangements afford minimal negotiation power to authors.
Julien Chouraqui, the legal director at the French publishing union (SNE), stated that these accords signify “progress” in addressing the challenges posed by advancements in AI technology.
The press is actively organizing efforts to contend with the evolving challenges presented by the rise of AI-generated content in the literary landscape.
Tech companies may have no choice but to compensate publishers adequately to enhance their products, especially as they find themselves increasingly depleting new material resources for their models.
“On the web, you find lots of licit and illicit content, and lots of pirated copies. That not only causes legal problems but also raises issues about the quality of the data,” Chouraqui emphasized, shedding light on the complex landscape of online content.
“If we are committed to developing a market on a virtuous basis, we must involve all the players,” he concluded, highlighting the necessity for collaboration across all stakeholders in this rapidly evolving industry.