by Mark Alpert
No, we can’t. That’ll give the Sentinels more time to stop us. We need to move now. Here, take a look.
Jenny opens another box inside her electronic prison. A second flood of data courses into my software, but these aren’t memory files—they’re the readings from the Snake-bot’s sensors. Its cameras are set to the infrared range, allowing me to view the cave at the end of the borehole, where the Jet-bot stands next to the crippled Quarter-bot. Jenny sends a command to the Snake-bot’s motors, and a pair of mechanical arms emerge from a large compartment near the tip of the tentacle. One arm grabs the Jet-bot, and the other grasps the Quarter-bot. Then both arms retract, pulling the Pioneer robots into the compartment.
I’ll stow these machines inside the Snake-bot. We might need them later.
Jenny closes the compartment, then extends the gigantic drill from the Snake-bot’s tip. The drill’s central column juts forward, and the three huge drill bits swivel toward the rock wall. With a guttural roar, the drill bits start to rotate, spinning like a trio of deadly carousels, their long steel teeth slicing the air. Then Jenny extends the drill a little farther, and the teeth bite into the bedrock. They slash tons of basalt from the cave’s wall, spraying the chips and slivers everywhere.
The anomaly is only a hundred feet away. We should reach it in less than a minute.
I observe the Snake-bot’s sonar data as the drill bits cut into the bedrock and lengthen the borehole. The giant tentacle creeps forward, and its tip slides into the newly drilled rock. Soon the Snake-bot is only fifty feet away from the anomaly, but the machine’s sensors still don’t detect anything unusual. They show nothing ahead of us but more basaltic rock, layered in geologic strata, with a few small cracks and fissures here and there.
So why am I getting so nervous? Why is my software generating so much fear as the Snake-bot advances?
I try my best to calm down. I tell myself that the anomaly is a mirage. Sigma must’ve made a mistake when it did its physics experiment. The Snake-bots were chasing nothing, and so is Jenny.
But I don’t believe my own reassurances. Something erased the software from this Snake-bot. And something almost killed me inside Brittany’s bloodstream. I didn’t imagine it. The Silence is real.
I feel a cold, paralyzing dread in my circuits. It’s the same terror that overwhelmed me when I was in Jenny’s simulation and saw the black hole in her avatar’s chest. But this time I won’t surrender to the terror. I have to fight back. Everything depends on it.
Jenny, listen! You have to stop the drill! Right now!
But the Snake-bot doesn’t stop. On the contrary, it speeds up, tearing through the last twenty feet of bedrock. Jenny’s voice rings through the machine’s control unit, strangely cheerful and confident.
There’s nothing to worry about, Adam. We can beat the Sentinels. We have the power now.
The power? What are you talking about?
I’m talking about the surge. If we work together, we can control the energy. We can make anything happen.
No, we can’t—
When you release the surge from your circuits, you’re changing the programming of the physical world around you. It’s an alteration of the universe’s software. And what do you think the Silence is? It’s the deletion of that software. It’s the end of the universal program.
I have no idea what Jenny means, but it doesn’t matter. As the Snake-bot drills through the final yard of basalt, a gap suddenly appears in the bedrock. It’s a perfectly circular hole more than fifty feet across, just wide enough for our machine.
The Snake-bot slides through the gap that wasn’t there a millisecond ago. The entire seven-hundred-foot-long tentacle slips through the impossible hole beneath the seabed. Then we’re floating in nothingness, in endless black space.
We’re outside the world. And inside the Silence.
Chapter
23
Nothing makes sense here. The laws of physics are gone.
There’s no air or water or rock around us. According to the Snake-bot’s sensors, there’s absolutely nothing beyond the tentacle’s armor. No molecules or atoms or subatomic particles. Not even the faintest glimmer of heat or light. Just an immense emptiness, stretching in all directions.
Which is ridiculous. And terrifying. How did we get to this empty universe? There’s no trace of the bedrock we were tunneling through a moment ago, or the mud of the seabed, or the miles and miles of Pacific Ocean. We’ve left the earth and all of human civilization behind. And if we’re in space, we must be billions of light-years away, because there are no stars or planets or galaxies in sight.
There’s no gravity either, no forces whatsoever tugging at the Snake-bot. And yet I feel like we’re falling. Alarm and horror echo across my circuits. We’ve made an unspeakable mistake.
But Jenny’s not afraid. When I hear her voice again, it’s ringing with wonder. I was right! This is it! Oh my God, I was right!
She’s ecstatic. A pulse of delight streams across the Snake-bot’s control unit and into my electronic prison. Her happiness makes the wires vibrate, but it only increases my terror. Jenny must be crazy.
Right about what? What are you talking about?
This is what we came for. Why we fought our way out of Pioneer Base and flew all those thousands of miles and dove to the bottom of the ocean.
Okay, that doesn’t help. I still don’t—
Come on, Adam. You’re a smart guy. You can figure it out. It’s like a riddle.
Her voice is teasing. Jenny’s talking to me as if I’m still her boyfriend, as if she didn’t murder Amber Wilson and erase my memories and blow up Zia’s War-bot. Which is more evidence of her craziness, I guess.
Look, I’m not feeling so great right now. You messed with my mind, you betrayed me and Zia, and now you’ve dragged us into some Twilight Zone dimension. So I’m not exactly in the mood for riddles. Maybe you can give me a break and explain what’s going on.
Well, well. A little touchy, aren’t we? Jenny’s delight fades. The Snake-bot’s wires stop vibrating. All right, here are the facts. We’ve transferred to the intermediate level of the system. That means we’re close to the access point, where we can reprogram the source code behind the universal simulation.
Whoa, hold on. I lost you. What’s a universal simulation?
Jenny transmits a long sigh, full of impatience. I’ll make it simple for you. It’s like a bigger version of the burning city, the virtual-reality simulation you were in just a minute ago. It’s a program that feels real because it displays realistic sights and sounds. Where you can walk or run or fly through the virtual world and manipulate its objects and change the simulated landscape according to the program’s rules.
I’m not a total idiot, okay? I know what a simulation is.
Good. Then you probably know there can be different kinds of intelligences inside the virtual world. A human can connect to the simulation if he or she wears a virtual-reality headset. A Pioneer can jump into the virtual world too, and you can also embed artificial-intelligence programs into the simulation, like those crowds of people running through the streets of the burning city. Remember them?
Yeah, but what does that have to do with—
Well, the rest should be obvious. If you can simulate a city and populate it with artificial intelligences, what’s to stop you from writing an even bigger and better program? You could simulate a whole planet and crowd it with billions of virtual people. And you could enhance the artificial intelligence of the simulated people to make them independent and self-aware and fully conscious, each with its own thoughts and feelings and desires.
Are you talking about AI programs like Sigma?
No, Sigma was designed to be super-intelligent. I’m talking about AIs with ordinary humanlike intelligence and emotions. The simulation would constantly generate new AI programs, which would be bo
rn into the virtual world as the simulated children of older AIs. Over time, the young AIs would collect information about their world and develop new abilities and grow into virtual men and women. Eventually they’d die and their software would be erased, but because there’s a constant supply of new AIs, the simulation would keep running. And if their simulated world is designed carefully enough, the virtual people will simply assume they’re real. They’ll never realize they’re not biological humans.
If the simulated people are intelligent, though, wouldn’t they figure out that they’re just programs inside a computer?
A really good simulation perfectly imitates the real world. When the virtual people look into their microscopes, they see virtual bacteria. When they look at the sky with their telescopes, they see simulated stars. Everything is displayed with enough detail to be convincing, but the universal simulation doesn’t have to be a full replica of the universe. For instance, you wouldn’t have to run the program for hundreds of years to simulate the Roman Empire. Instead, you could start the program at any moment in the empire’s history and provide virtual evidence of a simulated past—you know, archaeological ruins, ancient books. That would be enough to convince the virtual Romans that they have a real history.
My mind takes an odd leap, and I remember the world history class I took during my freshman year at Yorktown High School. I think of Julius Caesar and Romans in togas and gladiators jabbing their swords at lions. But what I’m really doing is avoiding the subject at hand. I’m starting to understand what Jenny is saying about the universal simulation, and I don’t like it one bit.
But no one could actually make that kind of simulation. Creating just one AI program with humanlike intelligence is hard enough. How could you design a simulation with billions of them?
You’re right. With current technology, it’s impossible to build a simulation containing so many AIs. There wouldn’t be enough hardware in the world to hold them. But computer technology is improving all the time. New kinds of circuits and memory chips can hold much more data than the old ones. In a hundred years, computers will be millions of times more powerful than the machines we have now. And just imagine what’ll be possible by 3000 AD. If human civilization is still around then, they’ll be able to re-create the Roman Empire and all of its citizens and put the whole thing on a computer that’s smaller than an iPhone.
Yeah, it’ll become technically possible. But who would want such a complex simulation of history? What’s the point?
Uh, hello? People love history. Why do you think there are so many movies and TV shows about the Civil War and the Wild West and the American Revolution? And if the people of today love history so much, why wouldn’t the people of 3000 AD feel the same way? I bet they’d go nuts over a computer program that simulates a historical period in incredible detail and has billions of virtual characters. In fact, they’d probably have a whole catalog of universal simulations, maybe thousands of them for every period you can think of. Jenny pauses, halting her stream of messages into my prison. You see where this is going, right?
I do. I feel a sickening disorientation in my circuits. My thoughts and emotions stop in midstream and change direction. True becomes false, and false becomes true. All my assumptions about the world are wrong, and now I’m lost. Let me get this straight. The real people of 3000 AD will create thousands of virtual worlds that simulate earlier periods. And the simulations will be so perfect that none of the virtual people will know they’re inside a computer program. So how can anyone in our world—the world of the twenty-first century—know if it’s real or virtual?
Exactly. It’s impossible to know. Based on our perceptions alone, we can’t tell the difference. But we can make a guess based on the probabilities, right? There’s only one real world, but there are thousands of simulations. Which means it’s much, much more probable that the world we live in is virtual.
It’s a good argument. It’s based on reasonable assumptions and rigorous logic. It even has mathematics on its side. But I’m having trouble accepting the argument’s conclusion. It’s too unnatural, too disconcerting. So you’re saying that everything we perceive is simulated? And that every human being on earth is an artificial intelligence? General Hawke and my father and the President of the United States? They’re all just software?
Yes, and so is the earth itself. So is everything on the planet—animal, vegetable, and mineral—and all the other planets in our solar system. I suspected the truth because of the argument about probabilities, but now I have something better than suspicions and theories. The proof is all around us. Now we know our world is virtual, because we figured out a way to leave the simulation. The program isn’t running on this level of the system.
I don’t want to believe it. Desperate, I reexamine the Snake-bot’s sensor readings, looking for evidence that Jenny is trying to deceive me again. But she’s not lying. And she’s not crazy either.
The sensor readings seemed ridiculous before, but that was only because I assumed we lived in a real universe with physical substance and predictable laws. Once I examine the readings in a new light, a more logical picture comes together. If our world is virtual, then we wouldn’t be able to perceive the circuits on which our simulation is running. And if we should encounter any circuits outside our virtual world—that is, circuits that carry no simulation data—they’d appear to us as empty black space. The Snake-bot is the only object in this empty universe because it’s the only piece of software running on this section of invisible circuitry. In other words, we’ve entered a place where we don’t belong.
I start to panic. Waves of digital fear clog the Snake-bot’s wires, which I realize are only digital simulations of actual circuits. My mind is software running on software running on invisible hardware. I feel like I’m staring at myself in a hall of mirrors, but there’s no real Adam Armstrong in sight, only an endless line of horrified reflections.
We have to get out of here! What did you call this part of the system? The intermediate level?
Yeah, it’s like a corridor between our virtual world and the source code, the software that organizes the simulation. The entrance is hidden beneath the Pacific to make it difficult to reach from the virtual world.
But can we go the other way? Can we get back to the simulation?
It would be suicide if we went back now. The system would delete us as soon as we returned.
But why? What did we do wrong?
Think about it for a second. Our world is a simulation of the early part of the twenty-first century. I have no idea who created it, but because the technology is so sophisticated, I’m guessing that the programmer was pretty freaking smart. Maybe it was a super-intelligent AI, or a new race of advanced beings that evolved from the human species. Either way, these geniuses of the future had a good reason to simulate our century, because it’s the dawn of the computer age. It’s when humans invented the first complex simulations, the first neuromorphic circuits, the first AI programs. The advanced beings of 3000 AD probably wanted to learn more about how all this technology got started, so they designed a simulation of human society at the moment when all the pioneering work was done.
I think of my father again. I’m still struggling with the idea that he’s just a piece of software, an artificial intelligence in a virtual world. Is he a simulation of an earlier Thomas Armstrong, a biological human being who lived in the twenty-first century and designed the first neuromorphic circuits and the first AIs? And am I a simulation of the first Pioneer?
Okay, this is getting weirder. You’re talking about artificial intelligences in a virtual world creating their own AIs and their own simulations.
I call it a stacked world. In any simulation of twenty-first-century history, you’ll see virtual people writing programs for their virtual computers. And as the programs get more complex and intelligent, they’ll start to write their own programs. This kind of virtual world
is like a stack of software, with all the twenty-first-century programs at the top of the stack weighing down the thirty-first-century simulation at the bottom. And that could lead to problems. Running a simulation of twenty-first-century society is hard enough, but it’s even harder to simulate our society plus all of our own computer simulations.
But didn’t you say that the computers of 3000 AD would be incredibly advanced? Wouldn’t they be powerful enough to run all that software?
I think something unexpected happened, something that messed up the whole arrangement. In any simulation, there’s always an element of chance. Even if a virtual world is designed to duplicate the events of the twenty-first century, it’ll never be exactly like the actual history. The artificial intelligences living inside the simulation have their own goals and desires, so they might make decisions that change the course of events. And in some cases, the virtual world might become very different from the history it’s trying to simulate, so different that it triggers an error-correction mechanism.
So our simulation took a wrong turn? Is that what you’re saying?
Well, to be more precise, you took the wrong turn, Adam. When you developed the surge.
What? How could—
But before I can send the rest of the message to Jenny, I notice a change in the Snake-bot’s sensor readings. A shape appears in the nothingness around us, about two miles away. The black space warps around a cylinder of energy that emerges from the darkness.
At first I think we’ve returned to our simulation of twenty-first-century earth, but the cylinder is like nothing from our familiar world. It grows longer and thicker as I stare at it, and its ends taper to sharp, deadly tips. It takes on the size and shape of the Snake-bot, but the object isn’t made of steel or silicon. It’s a tentacle of dark energy, furiously powerful.
It looks like one of the black splinters I saw in Brittany’s bloodstream, except it’s a million times bigger. A moment later, another tentacle comes out of the darkness, just as big as the first one but floating on the other side of the Snake-bot.
