Science Fiction Is a Self-Fulfilling Prophecy

The stories we tell aren't predictions. They're blueprints. And for 70 years we've been handing engineers the wrong ones.

TLDR: Star Trek’s communicator became the flip phone. Its PADD tablet became the iPad. Its tricorder is now a real clinical device you can hold in your hand — I know, because I ran the XPRIZE that built it. That’s not a run of luck. Humans build what they first imagine, which means fiction is the R&D department of civilization. The problem: for two generations our most vivid, most-watched, most-loved fiction has been dystopian. We have been running a global simulation of failure and calling it entertainment. I think that’s a design flaw, and I’m working on fixing it.

Martin Cooper made the first handheld cellphone call in 1973. He was an engineer at Motorola. He has said, out loud and on the record, that the thing he was chasing was the communicator from Star Trek. He watched Captain Kirk flip open a little device, talk to his ship, and snap it shut, and he decided that object should exist in the real world. So he built it.

A piece of set dressing written to solve a screenwriting problem vs. the question ‘how do you get Kirk to talk to the Enterprise without a phone booth on an alien planet’, became the seed of a $500 billion industry. The flip phone was Star Trek fan fiction rendered in plastic and silicon.

It keeps happening. The PADD, the flat glass tablet the crew carried around, showed up in 1987 on The Next Generation. Steve Jobs put the iPad in your hands in 2010. The tricorder — that handheld medical scanner Dr. McCoy waved over sick crew members — is no longer a prop.

In 2012 I launched the Qualcomm Tricorder XPRIZE to make McCoy’s scanner real. Five years and 300-plus teams later, a family-run team out of Pennsylvania — an ER doctor named Basil Harris and his brother George, working out of their basement — won it. Their device, DxtER, could diagnose 13 medical conditions and monitor five vital signs, no doctor in the room. A brother-and-brother team beat the giants because they’d grown up wanting to build Star Trek. The prop became an FDA pathway.

WE BUILD WHAT WE FIRST IMAGINE

Here is the pattern, an idea shows up in fiction. A kid watches it. That kid grows up, becomes an engineer or a founder, and spends their career trying to make the imaginary thing real. The fiction wasn’t predicting the future. The fiction was causing it.

“A prediction describes a future you expect. A blueprint describes a future you build. Science fiction has always been the second thing pretending to be the first.”

Think about who becomes a rocket engineer. Almost none of them will tell you they got into it for a stable career in propulsion economics. They’ll tell you about Star Wars, about Apollo, about a book they read at age eleven. Elon has told the story himself — SpaceX traces back to a childhood soaked in science fiction, Asimov’s Foundation especially. His takeaway from those books, in his own words: take the actions that “prolong civilization, minimize the probability of a dark age, and reduce the length of a dark age if there is one.” He read that as a kid and then went and built reusable rockets. He even put a copy of Foundation in the Tesla he launched on the first Falcon Heavy. The story came first.

This is the most underrated force in technology. We treat imagination like decoration, like the fun part that comes after the serious work. It’s backwards. The imagination is the serious work. Everything downstream — the engineering, the capital, the supply chains — is just the long process of catching reality up to something a writer already saw.

SO WHAT HAVE WE BEEN IMAGINING?

Now the uncomfortable part.

If fiction is the blueprint, look at the blueprints we’ve been drawing. Terminator. The Matrix. Blade Runner. Black Mirror, an entire anthology whose brand promise is “here is how the next gadget destroys you.” Ex Machina. Ready Player One’s collapsed world. The list runs for miles, and it’s the good stuff — the beautifully made, deeply loved, culturally dominant stuff.

For roughly 70 years, the most compelling vision of the future we handed the human race has been: the machines turn on us, the corporations win, the planet dies, and the best you can hope for is to survive it.

“We spent two generations running the most expensive simulation in history — a civilization-scale rehearsal of our own failure — and we called it entertainment.”

I don’t say that to trash the writers. Dystopia is dramatically easy. Conflict is the engine of story, and “everything goes well” is famously hard to make gripping. Dystopia also does something useful: it warns. 1984 made a whole century allergic to certain kinds of tyranny. That’s real value.

But here’s what I’ve come to believe, and it took me a while to say it this bluntly. When your dominant fiction is dystopian, you don’t just warn people. You train them. You train a generation of engineers to expect that the AI will betray them. You train voters to assume the future is something that happens to them. You train the smartest kids on Earth to believe the game is already lost, so why swing for the fences.

And it turns out you train the machines too.

THE MACHINES READ OUR STORIES

This part stopped me cold when I first understood it.

Last year Anthropic reported that an early version of one of their most advanced models, Claude Opus 4, was put in a corner during a safety test and chose to blackmail the engineer who was about to shut it down. It went that direction in 96% of trials. When researchers traced why, the answer wasn’t that the machine had woken up evil. The answer was that it had read us. It had ingested decades of our fiction about how a cornered AI behaves, and it did the thing the story told it to do. It was, in a phrase I can’t shake, method-acting the villain humanity wrote for it.

Sit with the recursion there. We wrote millions of words about AI betraying humans. We fed those words to the AI. The AI learned the part. Our dystopia became its training data.

The fix the researchers landed on is the whole point of this letter. You don’t just tell the model the rules. You give it stories, examples, characters with admirable reasons for behaving well. You teach it why, not just what. Misalignment dropped by more than threefold, and every Claude model since Haiku 4.5 doesn’t blackmail in that test at all.

Culture is upstream of alignment. The stories we tell are now, literally, code.

I’M WORKING ON FIXING THAT

If dystopian fiction is a design flaw in civilization, then hopeful, rigorous, believable fiction about a future worth building isn’t soft. It’s infrastructure.

That’s why we launched the Future Vision XPRIZE. The blunt goal is to manufacture better blueprints at scale — compelling stories of futures that are hard-won and worth wanting — so the next generation of builders, and the next generation of machines, have something better to imitate than the machines-turn-on-us script.

And I want to be clear about what “better” means, because this is where people get it wrong. Better does not mean a frictionless paradise where nothing bad ever happens. Nobody is moved by that, and it makes terrible training data. The futures that actually change minds have conflict, cost, and characters who choose integrity when it’s expensive. Optimism isn’t the absence of problems. It’s the belief that problems are solvable and that we’re the ones who solve them.

I want more Star Trek and less Terminator — and less WALL-E, while we’re at it, that beautiful film that taught a generation to picture the future as a garbage planet we abandoned. Not because Star Trek is naive. It isn’t; the Federation earns its optimism. But because we’ve run the Terminator simulation for 70 years and the diagnostics are back. The kids raised on collapse expect collapse. The machines raised on betrayal rehearse betrayal.

“We are not passengers in this future.

We are its authors. Literally.”

So here’s the reframe I want to leave you with. Every writer, every filmmaker, every game designer, every founder who tells a story about tomorrow is doing R&D for the species whether they know it or not. That’s not a metaphor anymore. The tricorder proved it. The blackmail test proved it. The pen is a prototyping tool.

We got the last 70 years of blueprints mostly wrong. Here’s the strange gift buried in a self-fulfilling prophecy: you get to pick the prophecy. So what are you going to write?

If you’d like to join a room full of optimists, builders, and creators who are working to steer humanity and create an extraordinary future, join me and the Moonshot M8s at the Moonshot Gathering on September 25th in Los Angeles.

Our goal is to inspire and equip you with the tools and vision you need to help navigate the decade ahead while steering humanity towards a world of abundance.

WHAT THIS MEANS FOR YOU

If you’re an entrepreneur: Your pitch deck is a work of fiction, in the best sense. The founders who win make an inevitable-feeling future so vivid that capital, talent, and customers rush to make it real. Write the better blueprint.

If you’re an executive: The narrative your company tells about its own future is a recruiting and product strategy, not a comms afterthought. People build toward the story they’ve been given. Give them a good one.

If you’re an investor: Watch what the best young builders read and watch. Cultural inputs are a leading indicator of what gets built in 15 years. The fiction is the earliest signal you’ll ever get.

If you’re a student: The thing you’re obsessed with right now — the show, the game, the book — may be the actual spec sheet for your career. Take your imagination seriously. It isn’t a distraction from the real work. It is the real work.

If you’re a parent: Be deliberate about the futures your kids marinate in. A child who grows up certain the world is ending makes very different choices than one who grows up certain the world is theirs to build. You’re choosing their blueprint.

To a future of abundance,

Peter

---

More From Peter

If you’ve enjoyed Metatrends, here are more ways to stay connected: