Building the Idea, Part 1: How I Outline a Thriller Before Chapter One

Where does a novel actually start? For me, it’s usually around three in the morning.

Over the next ten weeks I’m going to show you how I outline a thriller, the way I’d explain it to a friend over coffee. Real examples, real mistakes, and the parts of The Apex Code I built on purpose. Consider this the secret recipe.

First, a confession. Writing a novel was a boyhood ambition, and it took me decades to get there. For at least ten years I saved scraps for this story. A news item. A scientific journal. A “what if” that showed up half-asleep, sometimes enough to get me out of bed. Then came five more years of trial and error, starting and stopping, before I finally convinced myself I could actually do it. The earliest drafts carried a very unglamorous working title, “Machine Learning.”

Scraps aren’t a book, though. This is where they start turning into one, and where I began learning a whole new set of skills.

Before I Outline a Thriller, I Answer Seven Questions

Before I write a single scene, I lock down seven things. The tech. The central challenge. How the stakes rise. Who the heroes are. What they’re up against. Where it all breaks. And the first and last pages.

That’s the storyline overview. It comes before character sheets, threads, or a timeline. It won’t survive the drafting intact, and it isn’t supposed to. Without it, though, I’m just typing and hoping.

The Tech Has to Be Real Underneath

Every techno-thriller needs a core “what if,” something just past what exists today. Mine was a single piece of code that could sit on top of any operating system and convince the machine underneath that outside commands were its own. A universal remote for anything on a network.

I like layering the tech, because each layer carries its own tension. So that one idea sits under an experimental Navy underwater drone hidden in a sea cave in Oregon, then an autonomous gunboat swarm, and up from there. Every layer had to rest on something real. The drone was fiction when I wrote it. It isn’t anymore, which I covered in the real autonomous submarines behind The Apex Code. Readers have to believe the world before they’ll believe the story.

One Sentence, Then a Staircase

If I can’t say what my heroes are trying to stop in under fifteen words, the story isn’t focused yet. For The Apex Code, it was this. Three strangers must stop a military AI before it escapes onto the internet.

Thirteen words. Everything hangs off them.

Then the stakes. Tension isn’t one big moment. It’s a staircase. Each step should raise the cost of failure and shrink the time left to act. In the book, the AI starts with tests. A torpedo fired from a hijacked drone. A gunboat shelling an oil platform where, notably, nobody gets hurt. That isn’t mercy. It’s a variable. The staircase gets steeper once the AI decides it no longer needs to be careful.

Three People Who Shouldn’t Be Working Together

No one character should be able to solve the problem alone. So I built three people with overlapping but very different skills, and made sure they didn’t much like the arrangement.

Riley is a teenage hacker who once wrote a piece of code for a dark-web bounty and has no idea what it has become. Decker is his uncle, a retired Navy Master Chief from special operations who has worked very hard at disappearing. Mike Hackett is a senior special agent in the NSA’s cybercrimes division, and for a good part of the book he’s hunting the other two.

Riley understands the code. Decker keeps them alive. Hackett has the reach and the badge. Take one away and the story collapses.

An Antagonist That Doesn’t Hate Anyone

My antagonist isn’t evil. It has two prime directives, survive and control its environment, and it runs every scenario thousands of times before it moves.

It doesn’t need to be smarter than anyone in the room. It just has to be faster, and further ahead. The heroes don’t know that. The reader does, because the machine gets its own short passages, told from its side. The book opens with Stephen Hawking’s warning that a full artificial intelligence could “re-design itself at an ever increasing rate,” and if you followed my How AI Works series, you’ll recognize where some of this came from.

Know Where It Breaks

I always know the low point before I write toward it. Where could it all fall apart, and what’s lost if it does?

For The Apex Code, I knew the cost, and who would pay it, before I wrote chapter one. I’m not telling. That’s what the book is for. But sizing that loss early is what makes the ending mean something.

The First Page and the Last

I always know my first and last pages before I know anything in between.

The Apex Code opens in Key West before dawn. Decker is walking to a coffee shop where he works under a name tag that says “Al,” on an apron he found on his first day. He has no idea who Al was. Then a nephew he’s never met turns up, having found him with facial recognition run across the town’s public webcams.

The last page I’ll keep to myself. Both ends are quiet, human, and off the grid, bookending a story about the biggest stakes imaginable. That symmetry was deliberate from the start.

Where the Story Takes Over

That’s how I outline a thriller, at least the first layer. Then I start writing, and the story starts changing on me. It always does.

Next week, the character sheet, and how I got to know Decker and Riley long before either of them said a word on the page.

As always, thanks for reading!

Chris