And so we come again—as we will many, many times in the future—to the question of artificial intelligence.

I read an article recently in The New Yorker by a Princeton professor called D. Graham Burnett who wrote rather richly about pressing students to use AI for something more than just, you know, writing their papers.

Burnett writes, “I’m a book-reading, book-writing human—trained in a near-monastic devotion to canonical scholarship across the disciplines of history, philosophy, art, and literature. I’ve done this work for more than thirty years. And already the thousands of academic books lining my offices are beginning to feel like archeological artifacts. Why turn to them to answer a question? They are so oddly inefficient, so quirky in the paths they take through their material.”

I was already on that route. In my old sea captain’s house, I had an entire room with floor-to-ceiling bookcases, all of them filled with books. When I moved to my cottage, it was clear I had to downsize my library—but by that time I’d begun collecting more titles on my e-reader than in my hands. I kept some physical books—favorites, ones with some value or use to me, gifts I couldn’t part with—and donated the rest of it. My (smaller, fewer) bookcases now are still full, but Burnett is right: I rarely consult them, when I have the Internet at my fingers.

It's not the same, of course. Growing up in a household where learning was honored, I spent many, many hours with an encyclopedia in my hands, having gone to look up one thing and mesmerized into reading much, much more. There was a magic of discovery there that Wikipedia just can’t touch.

And honestly, if you had told me that I would one day own a tablet object that contains thousands of books, I’d have definitely thought you crazy.

That was crossing a threshold of sorts, wasn’t it? Learning about a completely revolutionary way of doing something, putting up a bit of resistance, and then seeing the richness it brought into my life, making it part of me.

We’re at a new threshold now, not with ebooks, but with artificial intelligence. And as with all liminal spaces, we’re largely uncomfortable about standing on the threshold. I think that goes for almost everyone; but it adds a certain layer of complexity for those whose livelihoods will most likely be affected.

Me, for example. I don’t have a Plan B. I have to echo the wonderful James Keelaghan, who said, “I’m a folksinger-songwriter. My retirement plan is death.” Yep.

Burnett’s question goes to the heart of what many of us are wondering these days: if this robot can write these books as well as—or even better than—us, then what are we doing here? The results of these LLMs are becoming more impressive every day… and it’s becoming more and more difficult to distinguish their output from what people write. And once you’ve asked that question, you can’t un-ask it. That horse is never getting back into that stable.

Ludwig Wittgenstein says that we form language as part of human activity, and that we then shape that activity through language. But both parts are necessary; and, so far, AI can’t manage the “human activity” part of the equation.

While technology advances—and few writers other than Michael Crichton are able to speculate what’s next—there is generally a sense of movement toward something that expands our knowledge of and participation in the world. We learn language to deepen and broaden that participation. And that process can’t really be experienced by machines, no matter how advanced.

If I write a book so that you can have that book in your hands, then AI can obviously replace me. But if I write the book in order to introduce you to ideas, to arouse feelings, to change the activity that you do in the world, then AI cannot replace me. It can give you another version of me that might be okay, better in some ways, worse in others—but not the same.

And for some that won’t matter. Many of us are skeptical of the current human experience, since it involves a chaotic world where emotions play as large a part as ideas and knowledge. For many of us, doing away with that “messiness” might be a worthy goal.

For now, I’m going to keep writing. From my experiences. From my feelings. From—dare I say it?—my heart.

Even if ten or twenty years from now a ChatGPT creates a Sydney Riley, or a Martine LeDuc, or an Abbie Bradford, or any of the other characters we create in our heads, we’ll have had the experience, once upon a time, of making magic.

And I’m not sure that will ever come from a machine.

Real or Robot? Who Can Tell?

Why We Write in Dark Times