Sylvia Plath committed suicide a month after her only novel, The Bell Jar, was published in the U.K. We can assume the vast majority of readers were aware of that fact or learned at some time after reading the book, and certainly knowing that the story was reflective of the author’s own struggle with depression colored their reading experience and even, I’d argue, their comprehension of the author’s words. This added dimension—the visceral connection with the author’s life experiences—will not exist for AI-generated fiction, nor non-fiction for that matter.
I’m on the side of those who believe AI will be able to produce rich, engaging, emotive work, but I see it as two-dimensional for that reason. Much will depend on the reader, the topic, the genre, and other variables, but I think it’s likely readers will prefer the feeling of enrichment that comes with knowing a human struggled to produce the work, and they’ll often want to know more about what motivated the author to spend months or years devoted to its creation.
Due to this added dimension, I feel confident human-penned literature will remain on a higher plane than AI work. My fear is, however, that very few writers will make a living at it because AI work will overwhelm publishers and perhaps even bring them down. (Why pay for a book when you can ask your favorite LLM to write you a perfectly appealing novel based on everything it knows about you and your proclivities, ambitions, etc.?) I hope in ten years we’ll still be debating whether humans write better books than LLMs, but if no one can make a living as a writer, it will greatly reduce the possibility humans can stay ahead in this competition.
Sylvia Plath committed suicide a month after her only novel, The Bell Jar, was published in the U.K. We can assume the vast majority of readers were aware of that fact or learned at some time after reading the book, and certainly knowing that the story was reflective of the author’s own struggle with depression colored their reading experience and even, I’d argue, their comprehension of the author’s words. This added dimension—the visceral connection with the author’s life experiences—will not exist for AI-generated fiction, nor non-fiction for that matter.
I’m on the side of those who believe AI will be able to produce rich, engaging, emotive work, but I see it as two-dimensional for that reason. Much will depend on the reader, the topic, the genre, and other variables, but I think it’s likely readers will prefer the feeling of enrichment that comes with knowing a human struggled to produce the work, and they’ll often want to know more about what motivated the author to spend months or years devoted to its creation.
Due to this added dimension, I feel confident human-penned literature will remain on a higher plane than AI work. My fear is, however, that very few writers will make a living at it because AI work will overwhelm publishers and perhaps even bring them down. (Why pay for a book when you can ask your favorite LLM to write you a perfectly appealing novel based on everything it knows about you and your proclivities, ambitions, etc.?) I hope in ten years we’ll still be debating whether humans write better books than LLMs, but if no one can make a living as a writer, it will greatly reduce the possibility humans can stay ahead in this competition.