Source: in-cyprus.philenews.com
Olga Tokarczuk grieves the death of the “old way” of writing — but thinks AI could be a boon for fiction
What makes a literary novel appealing to readers — not all readers, obviously, since tastes vary wildly? The subject matter and how the author handles it. The distinctiveness of their voice. Sometimes, frankly, a well-funded marketing campaign. There are writers whose books are declared bestsellers before a single word has been written, though their actual worth is debatable — these are the ones I personally avoid. And then there are genuinely fine writers whose work never quite found the audience it deserved. But if the goal is to give readers pleasure, why is there such an uproar when a writer uses AI to write better?
A few days ago, Polish author Olga Tokarczuk — Nobel laureate in Literature, 2018 — admitted with disarming frankness (or was it naivety?) that she uses an advanced AI model not just for research and information-gathering, but for plotting and building characters. “Often, I simply ask the machine, ‘darling, how could we develop this beautifully?’” she said in an interview. She acknowledged she’s aware of the algorithms’ limitations, but believes AI can be a tremendous asset for fiction. It seems she’s not alone — plenty of writers are apparently doing the same thing quietly, and keeping very quiet about it.
Tokarczuk and her publisher were quick to clarify that no book has actually been written with AI. But in that same interview she said something far more interesting — about how literature itself is changing. She announced that the book she is currently writing will be her last, because she believes readers no longer have the appetite for complex literary works. She also said it saddens her that the “old way” has run its course.
Has it, though? And is that good or bad — for literature, for writers, for readers? What happens to prizes like the Nobel? Many are already calling for Tokarczuk’s to be revoked, even though AI as we know it didn’t exist before 2018. And what does the future hold? Will we one day be handing awards to the best chatbots and the cleverest prompts?
We’ve known for years about writers who churn out doorstop novels at record speed — yes, their publishers slot them straight into the bestseller lists. Before AI existed, you could read one of their books and feel you’d read all of them, including the ones they hadn’t written yet. And yet there was always — and still is — a devoted readership that compulsively snapped up each new supposed book the moment it appeared, making it a genuine bestseller through sheer purchasing power. But what pleasure is there in reading something that feels oddly familiar, as though you’ve been here before? The consumer of digestible, formulaic, saccharine output is a mystery unto themselves — and they were at it long before anyone invented a machine to do the same job.
Incidentally, calling it “intelligence” is rather generous. A machine has not a shred of it — at least not yet — because it lacks the capacity for original creation. It simply hoovers up fragments of existing information and reassembles them into whatever new shape it’s asked for. But then again — don’t we all do exactly that? Aren’t we, too, information-gatherers, storing texts and news and ideas in our memories, then synthesising something — a novel, an article, a study — from things we’ve already encountered, things we absorbed somewhere along the way and now consider our own? And how new is that, really? In what fundamental way does an AI-generated work differ? I personally refuse to use it — but why shouldn’t it be considered as legitimate as any other, more “natural” act of creation?
Artificial intelligence — I use the term for want of a better one — is already embedded across an enormous range of human activity, not just the arts. Some applications are considered acceptable; others, less so. The fears and dangers surrounding its growing use are, on the whole, reasonable — we’ve been rehearsing them through science fiction for a century. Perhaps the moment has come for them to become real. Can we harness it for good, turn it to our advantage, and rule out its misuse — perhaps by using the very same technology to establish inviolable ethical guardrails? The devil’s advocate in my head whispers that nothing is truly inviolable. But I think it’s worth trying, given that we don’t have many alternatives.
One of the great fears around widespread AI use is the sidelining of human beings — the same fear that gripped people during the Industrial Revolution two and a half centuries ago. Technology is welcome when it liberates workers, as long as nobody ends up unable to put food on the table. But then look at our leaders — in the world at large, and on our own little island. After last Sunday’s elections, ask yourself honestly: would you prefer a parliament run by artificial intelligence over a parliament — and quite possibly a government — of no intelligence whatsoever, like this one, or the ones before it? I don’t know if that makes me a nihilist. I’m certainly disillusioned, and I’m certainly pessimistic. Someone once said that stupidity is unbeatable — and I’m referring specifically to the stupidity of the average voter. Does AI carry any trace of stupidity? Probably about as much as the people who build it and use it.