ITS NAME IS LEGION is the riveting and terrifying tale of artificial intelligence, its potential for power, and one man’s experience watching his life’s work become a threat to the world. What was the inspiration for the setup of your story?

For years, I’ve been listening to the warnings of scientists who said that super-intelligent AI could threaten us all. When Stephen Hawking said, “The development of full artificial intelligence could spell the end of the human race,” he got my attention.

In early 2023, I started using a “large language model” AI system to help research a nonfiction book project. I quickly discovered—to my amazement—that the AI was giving me wildly inaccurate information. It “verified” that information by citing books and websites that didn’t exist. It gave me “quotations” that were invented and nonexistent. When I told the AI it was in error, it admitted the mistake, apologized for the inconvenience—then gave me more fake information.

I soon discovered that AI engineers are well aware that AI systems lie. But they don’t call AI falsehoods “lies”—they call them “hallucinations.” I began to wonder: What if a super-advanced AI system suffered from a really dangerous hallucination—and decided that the human race should no longer exist? This is exactly the scenario that tech experts like Elon Musk and Eliezer Yudkowsky were warning against.

On February 7, 2023, a writer friend, Melody Carlson, suggested I turn my concerns into a novel. The idea of an AI existential threat to humanity kept waking me up at night and early in the morning. Finally, on February 19, I started writing the first chapter. I finished a complete draft five weeks later, on March 25. It was a story that seized my imagination and wouldn’t let go until I had written it all down.

I found the science in the novel to be well-developed. What kind of research did you do to make sure you got it all right?

I had been following developments in AI for a few years. Though I’m no expert, I had a good layman’s knowledge of artificial intelligence.

For me, the story comes first, research last. I started writing without researching, without outlining. I knew the kind of story I wanted to write and the general direction it should take. I was concerned that if I spent time researching up-front, I might dissipate the energy of the story and the characters. Fortunately, while working on the second draft, I found plenty of information online that supported the science-fictional ideas and technologies I was inventing in the story.

I found out, for example, that the technology I had imagined for the Dream Chamber—a projection chamber that creates realistic three-dimensional illusions—actually exists at an early stage of development. My intuition told me that such a technology was possible. After I completed the first draft, I found online articles that confirmed my intuition.

What was your favorite scene in this story?

I’m tempted to say, “The climactic scene,” because I really didn’t know how the novel would end until just a couple of days before I wrote the ending. I was very relieved when the conclusion of the novel finally came to me.

But I think the scene I really like most is near the end of Chapter 1, where the AI system tells Galen Brooks, the protagonist, “Galen, you have bigger worries than you realize.” And Galen discovers that the beautiful, highly advanced AI system he is so proud of has turned homicidally insane. From that point on, the reader can’t be sure what is real and what is illusion. To be candid, the author was never sure, either.

What is the next book you are working on, and when will it be available?

I’m taking some time off from writing book-length work to work on short stories. I’ve really wanted to produce a body of short works. Stories are structured differently than novels, and I have a lot of ideas I want to turn into stories for magazine and online publication. When they sell, I’ll post the news on my website.

Author Links: Goodreads | Twitter | Website | Amazon