Toys Of Ultimate Power
In 1997, I began a strange journey. It started in a writer workshop in Seattle by total accident and it has changed both the course of my life and how I think about the world. I was attending Clarion West, an illustrious bootcamp for aspiring science fiction writers.
Clarion West is a marvelous institution. If you can win a place, you have six weeks to discover what it means to be an author. Each week, some different award-winning luminary becomes your mentor, reading your work, helping you hone your skills. Or rather, I should say, that’s what happens for five of those weeks. There’s one week that’s different (or it was in 1997), where representatives of a major US science fiction publisher serve as your mentors instead. You are encouraged to produce a novel outline. As an exercise, you understand, not as a pitch to the publishers. However, everyone knows that if the publishers like your pitch, magic can happen.
My pitch was for a hard science fiction crime novel that riffed hard off my prior experience as an AI researcher. It involved espionage, uploaded consciousness, Egyptian mythology, and noir detectives. I thought it was pretty cool.
About halfway through the critique session for my outline, a friend of mine in the group offered the following thought.
“It’s a great premise, Alex,” she said. “It’s a shame it could never happen in real life.”
Needless to say, I wasn’t thrilled by this assessment under the circumstances, even coming from someone I admired. I asked her what she meant.
“Consciousness can’t run on machines,” she told me. “Roger Penrose has proved it. It’s all in his book, The Emperor’s New Mind.”
“Bollocks,” I thought, but did not say.
It didn’t seem right to debate the point with her during a crit-session. That would have been counter to the spirit of the workshop. But I filed the remark away and resolved to read this book that had supposedly upended the last thirty years of AI research.
I read it that summer, from cover to cover. In his book, Penrose makes the claim that human consciousness can’t be computed because the mind relies on processes in nature that are themselves non-computable. To make his point, he delivers a tour of all the stranger parts of relativity and quantum mechanics.
The book had a remarkable effect on me but not the one Penrose intended. He convinced me that he was exactly wrong. Not only was the human mind computable, but every weird physical phenomenon he outlined was computable too. Nature, I came to believe, could be described using nothing more than a discrete process proceeding via a system of simple, inexorable rules, up to and including all the surreal phenomena of quantum mechanics.
For whatever reason, this idea took hold of me. And over the years that followed, I found myself embarking on a quixotic quest: to prove Penrose’s error. I wanted to take every important physical symmetry that informs how our universe operates and reproduce it in discretized, digital form. I wanted to model everything. Not all at once, you understand, but piecemeal – one symmetry at a time in a series of concise but irrefutable digital toy models.
For all its lunatic ambition, this quest served me well. I succeeded and proved my point, even though the results I came up with are never likely to be useful to a working physicist. Along the way, I found myself learning a large enough body of physics as an adult to enable be to get occasional jobs as a visiting quantum gravity researcher. But my quest did more than that. By teaching myself how to think and operate like a scientist, I started to get other scientific jobs. Physicists, as it turned out, were usually disdainful of algorithmic models. (An attitude I still find wildly unimpressive.) However, biologists, psychologists, sociologists, and network scientists definitely were interested. And the skills I acquired decrypting the mysteries of quantum mechanics turned out to be remarkably useful when modeling gossip propagation or the distributions of African termite mounds.
Because I have at no point been a career scientist, I have been at liberty to wander and collaborate on whatever projects I like. While it has never been lucrative, this journey has given me an opportunity to study a large number of scientific disicplines, if never in depth. It led me to a position a Princeton University and recently a collaboration with the SETI Institute creating artifical life from scratch. It has convinced me that the use of simulation represents a new and incredibly powerful way to extend human reasoning.
I am not talking about A New Kind of Science as Stephen Wolfram puts it, though I am more sympathetic to some of his ideas than many I encounter. But rather that simulation represents a new form of augmented intuition. In his lovely book How Not To Be wrong, Jordan Ellenberg describes mathematics as a tool to extend the reach of human thought. I believe that the ability to build simulation confers that power also, while not augmenting it in the same direction. The intuition that simulation grants you is at once less incisive than mathemetical thought, but it is faster, broader, and much easier to learn. It is a kind of extended mental superpower that anyone who can write a page of code can acquire if they are so inclined.
In this blog, I want to outline what I think simulation can do for you via a series of interactive examples. I don’t want to just convince you, dear reader, that this superpower exists. I want to share it. I want to reveal it by making little toy models of reality. In doing so, I hope to grant you the ability to understand human nature, comprehend the universe, and even occasionally see the future. It is my hope that along the way you will also be entertained.
So without any more preamble, let us be off. The simulated universe awaits. To finity, and beyond!