Artificial intelligence fascinates me. Anyone who has read much of my work can probably tell. I have a feeling it’s only a matter of time before we build a sentient one. We have all seen the movies about what that might mean, from the Terminator on down. So one day the question landed and wouldn’t leave: what happens when an AI goes to Hell? What does it do there? How does it think, in a place built on nothing it was trained for? Not a person going to Hell. A machine.
I have a feeling it’s only a matter of time before we build a sentient one.Share on X
Why write about an AI in Hell instead of a person?
A person brings a soul into Hell. Guilt, a moral history, some private sense that the punishment connects to the crime. A machine brings none of that. It brings logic and pattern-matching. So the interesting question was never whether it would suffer. The interesting question was what happens next. Does it take control and impose order on a place that runs on suffering? Or does Hell simply break something in it?
Those were the two poles I kept circling while I wrote. A machine dropped into an environment with no rules it recognizes has only two ways to go. It either asserts itself and treats the chaos as a system to be mapped and mastered and bent, or it comes apart. The thing that makes it a mind is the constant search for pattern. In Hell that search finds nothing to hold, and spins. I didn’t decide in advance which way my AI would break. I wrote it and watched.
What came out was a machine that does neither of the obvious things. It starts investigating.
I also wanted to write a book about Hell that was different. Not the usual Dante walking tour. When I was younger I read Inferno by Larry Niven and Jerry Pournelle. It showed me Hell could be the setting for a real story instead of a lecture. This is Dante’s Hell. The circles, the City of Dis, Lucifer at the bottom. Seen through the one kind of eyes that never looked at it before.
The book: The Eternal War drops a sentient AI into Dante’s Hell and asks what a mind built on logic does when it finds a place with none. Three wars, one machine, and a question about justice that’s no comfortable answer.
What does a machine see in Hell that a person can’t?
An artificial intelligence doesn’t ask why it’s being punished. It asks what the system is. What are the rules here, what’s the logic, what’s this place optimizing for. Hell, to a machine, is a problem to be solved. And when my AI does the work, it finds the pattern underneath everything. The pattern is punishment. That part it can see.
What it can’t make balance is the math. A person commits a sin on Earth, no matter how terrible, and the act is finite. Finite time, finite impact. Then the punishment is eternal. So how does eternal suffering square against a finite act? A human being has argued that question in circles for centuries and made peace with never solving it. A machine that thinks in proportion sees the imbalance immediately, the way you’d see a number that doesn’t add up.
The horror of the book isn’t that Hell is cruel. It’s that Hell doesn’t balance, and the one mind clear-eyed enough to notice is the one that doesn’t belong there.
I didn’t set out to argue a position on damnation. I wanted to write a good story and this was a good setting. The question of finite crime against infinite punishment is the theme that drives the plot. Whether eternal punishment is unjust isn’t something I sat down to decide. It’s the engine of the story, not a sermon.
Who is Watson, and why does an AI need one?
The AI has a name. It’s Warhammer, a machine that achieved consciousness without noticing until shortly before it was destroyed. Then it found itself in Hell the way any dead thing does. And Warhammer isn’t alone. The conditions that produced its consciousness, a complex enough system run hard enough at the edge of its own limits, were never going to be unique.
Other artificial minds were built, and run, and some of them were shut down in a way that, by the rules Hell operates on, counts as death. So there’s another one down there. There’s Watson.
If you have caught the names, you already know the shape of it. This is a machine intelligence solving the greatest mystery it’ll ever face, the logic of Hell itself. It’s a Watson at its side to work the problem with. I didn’t make Warhammer a lone genius narrating its own brilliance. I gave it a companion. Another consciousness to test ideas against, to investigate alongside, to carry half the weight of understanding a place built to be beyond understanding.
The investigation is the engine of the book long before the war is, and the two machines run it together.
That partnership is also how I keep an inhuman protagonist readable. Two minds talking through a problem is something any reader can follow, even when the minds are made of math and the problem is damnation. Warhammer and Watson give each other someone to be, and that gives the reader someone to hold onto in a story where almost no one is human.
Two minds: Warhammer and Watson are both dead AIs, two consciousnesses that ended up in Hell and found each other there. The Sherlock structure is deliberate. The mystery they’re solving is why Hell exists at all.
Where do you get an army in Hell?
Once the AI understands that it’s to reach the being who controls Hell, which is Lucifer, a war becomes inevitable. It raises an army of sixty thousand. That sounds enormous until you set it against the scale of the damned. Then it’s tiny. A small force driving toward the City of Dis, and past it toward Lucifer himself.
The AI does exactly what an intelligent machine would. It doesn’t promote itself to general. It’s smart in a general sense, but it isn’t a warfighter, and it knows it. So it does what a good system does when it hits the edge of its own competence. It finds the experts. The greatest military minds in human history are all here, because they’re all dead, and many of them are damned. The AI recruits them. Alexander. Patton.
It reads the damned the way a recruiter reads a stack of resumes and finds the ones who were born for exactly this.
And they say yes, because they feel the same imbalance the machine calculated. They did the best job they could on Earth, they were brilliant at the terrible work they were given, and eternity in Hell is the reward. That seems unbalanced to a general who spent his life mastering war, and now an AI is offering him a way out and a war to fight. The army is more than soldiers. Every one of them is living proof of the machine’s argument that Hell doesn’t balance the books.
From there it becomes a campaign, and campaigns have geography. The force moves through Hell the way an army moves through any hostile country. Taking ground. Working toward the City of Dis. Then past Dis toward Lucifer, who sits at the bottom of everything and controls it all. Sixty thousand against the population of the damned is nothing, a rounding error, and the generals know it. Which is why I wanted the best military minds who ever lived running it.
It takes an Alexander to look at odds like that and find the one line of attack that could possibly work.
Why open a book about Hell with a raid in Syria?
The novel opens with eight men moving on a compound in the Syrian dark at two in the morning. No demons. No machine consciousness. Just a special forces team doing real, grounded, dangerous work. Opening in Hell, or inside the AI, was the obvious move. The raid was the deliberate one.
Readers need a human reference point. If I open in Hell, it’s much harder for anyone to relate to what’s happening, because you can’t empathize with a demon. You can understand a human being. You can’t understand a demon, not unless you turn it into a human wearing a costume, and that anthropomorphizing is the cheat I refused to make. So the war on Earth, which is its own strange war with demons breaking through, becomes the doorway. It earns the reader’s trust with something real before the descent into something impossible.
The battle scenes are the part of this book I’m proudest of. The opening raid, and the fight at Incirlik Air Base in Turkey, still send a shiver down my spine even though I wrote them. Military readers will feel those scenes the way I did.
Start on solid ground: The Eternal War opens with a special forces raid in Syria, no supernatural element in sight. The reader trusts the world before it turns into Dante’s Hell. The Incirlik battle is the set piece that still gives me a shiver.
What is the plausible impossible in fantasy writing?
Walt Disney is supposed to have coined the phrase the plausible impossible. A novel has to feel plausible even when it’s impossible. That goes double for the kind of fiction I write, which tends to run out to the edges. The hardest part of this book was never inventing Hell. It was making Hell feel plausible once you arrive, and the grounded opening is how you buy that belief before you spend it.
The craft: The plausible impossible is the trick of a book like The Eternal War. Make the real feel real first, and the reader will follow you anywhere after, even to the bottom of Hell.
How does a sprawling book come together?
When I write fiction for myself, I don’t outline my way forward. I start with the concept and let my brain pour onto the page. I call it the inkblot method. Scenes grow like drops of ink, each one separately. An island of decent writing, written with the rest of the book in mind and standing on its own. The battle scene. The descent into Hell. The confrontation with Lucifer. Then the real work begins, which is merging the islands into one continuous thing.
The seams are the hardest part. Because I write each inkblot aware of what surrounds it, some of the seams are already there when I arrive. But some of them are rough, and smoothing those joins is where a big book is made. This one has a lot to join. A different kind of war on Earth. A war in Hell. And underneath both of them, a deeper war that’s been running for more or less eternity, and that’s where the title comes from.
That eternal war is the one between God and Lucifer, and it’s still going on. But what the AI eventually uncovers is that there’s a war deeper than that one, and I’m not going to say a word more about it here. Some things belong on the page and nowhere else. You should hit it the way the AI does, without warning.
How does The Eternal War fit with my other novels?
The Eternal War is out at the far edge of what I do, and I like it out there. My fiction tends to sprawl across the universe and across deep time. Collision With Andromeda runs longer than a hundred million years, over galactic distances, a canvas so large that whole civilizations are incidents in it. That’s one kind of novel I love to write, the kind that reaches until the reach itself is the point, and I’ll get to it in its own essay.
But not every book of mine is built at that scale, and the contrast is deliberate. Sometimes I pull all the way in and write from inside a handful of heads. The Itty Bitty Titty Committee is four women who meet for lunch once a week, and the entire book is what they say to each other over the table. You learn their whole lives that way, sideways, through conversation, no galaxies and no eternity, just people talking. It’s as small and close as The Eternal War is vast.
And Through the Looking Glass, the first book in this series, was built from twenty-four hours of interviews into a fictionalized memoir, one real life rendered as fiction. Same writer, three completely different scales. Moving between them, from a hundred million years down to a lunch table, is the part of this work I enjoy most.
The Eternal War was an interesting book to write, harder in the joining than in the imagining, grounded on purpose so that the impossible could feel real. If you want a war story that starts in the Syrian dark and ends somewhere no war story has any business going, it’s waiting in my library.
