Well, not really. Or at least not in the way that you might think: I’m definitely not one of those scammy side hustlers sending ChatGPT-generated concoctions to award winning science fiction magazines.

But the novel I’m working on now, Pacifica, begins each of its 74 chapters with an epigraph. Much like the computer game Civilization, each chapter is named after one of the technologies that have made modern humanity possible. And, much like Civilization, each technology is accompanied by an apposite quote. Leonard Nimoy was the gold standard narrator for those quotes in Civilization IV (though Sean Bean has his moments in Civilization VI).

One of the most fun parts of drafting Pacifica has been finding the right quotes for each chapter. I picked from books and poems that I love (as well as a few books that I hated) to put together what I imagined as a kind of collage or mosaic of human knowledge. I imagined the task as something like a literary version of the cover of Sgt. Pepper’s Lonely Hearts Club Band, where The Beatles assembled a photo-collage crowd of their favorite thinkers and artists and goofball influences.

Many technologies were easy to find quotes for. Especially for early technologies like pottery, masonry, and currency, there are a thousand great writers who had something pithy to say. Mostly I would page through books in my office, or CTRL-F through digitized books in Archive.org, to find quotes that spoke to the technology in question and also, hopefully, to the action of the chapter. Sometimes I had to draw the connections myself, in which case the quote turned into something of a writing prompt; other times the quote fit the chapter in deep and unexpected ways that I couldn’t have engineered if I tried.

Some of the later technologies were much harder: for instance, no one from Homer to Virginia Woolf seems to have much to say about the superconductor. Who could I quote for a tech like that?

It just so happened that by the time I got to the superconductor chapter of the book, everybody was talking about ChatGPT. At my college, the discussion revolves entirely around students’ using ChatGPT to plagiarize their essays, an issue which seems to me as trivial, in the grand scheme of dangers that ChatGPT represents, as the crew of the Titanic arguing about a shortage of urinal cakes in the men’s rooms of the Saloon Deck.

So I asked ChatGPT to find me some quotes about superconducting. It suggested some quotes from Larry Niven’s Ringworld and Niven’s and Jerry Pournelle’s The Mote in God’s Eye. They weren’t bad references, exactly–those books do mention superconducting–but none of them resonated with me. So I asked about Arthur C. Clarke, a fave of mine: surely, I thought, Clarke must have written somewhere about superconducting.

According to ChatGPT, Clarke has written about superconducting: of the two references ChatGPT gave me, the one which jumped out at me was this one: “Clarke’s short story ‘The Ultimate Melody,’ published in 1957, briefly mentions the use of superconducting materials in the construction of a futuristic musical instrument called the ‘ultimate melody.'” Now that’s a resonant quote–that would work perfectly for Pacifica!

So I looked up the story and read it (like 90% of Clarke’s short fiction, I had never read it before). Here’s the thing, though: there’s absolutely nothing about superconducting in that story! (For that matter, the futuristic musical instrument is called “Ludwig;” the ultimate melody was the ideal music the instrument was designed to find).

And here’s the other thing, which I discovered later: Arthur C. Clarke did write a short story, called “Crusade,” in which superconductivity is a central plot point. ChatGPT didn’t think to mention it (because ChatGPT doesn’t think yet). I tracked that story down with a simple DuckDuckGo search for “Arthur C. Clarke superconducting.” It’s an excellent story, by the way–very Arthur C. Clarke. And that story had the perfect quote, which fits both Pacifica and the life I feel I am living lately: “It was a computer’s paradise. No world could have been more hostile to life.“

So, for now, I agree with John Scalzi’s excellent assessment: “If you want a fast, infinite generator of competently-assembled bullshit, AI is your go-to source. For anything else, you still need a human.” That’s all changing, and changing faster than I would like, but I’m relieved to know that I’m still smarter than a computer for the next year or maybe two.