by David Toomey
Suppose, however, at some point in the future simulated organisms become both living and sentient. They might regard the computer simulation as their universe, and they would understand it to be something like our own traditional universe, in which laws are the same throughout. Tegmark’s conception of reality as a mathematical structure quite obviously applies here. Computer simulations are, fundamentally, a series of mathematical manipulations that represent that computer’s state. And insofar as these simulations are universes, their programs are their theories of everything.
AN UNSETTLING QUESTION
Ideas that one’s existence might be nothing but the dream or imagining of another have appeared throughout literature. They are also well-trod territory in philosophy, particularly in epistemology, the field of study that asks, among other questions, how we can be sure what we know. Computer simulations, and the probability of increasingly detailed and realistic simulations, have provided a new framework for that question.
Computer power is increasing rapidly, and will likely continue to increase. If quantum computing is ever realized, processing speed will be exponentially faster. The fascination in creating simulated worlds is enticing and, it seems reasonable to expect, will continue to entice, especially as those worlds grow ever vaster in scope and richer in detail.
For the moment, then, let’s allow our imaginations free reign. Members of a civilization with technology millennia in advance of ours might be able to create simulated universes, and assuming no technical obstacle, they might also be able to create inhabitants who are sentient and self-aware. The scenario of computer-generated alternate realities is common in science fiction, and it invites a question. To borrow the nearest cultural referent, how do we know we’re not living in the Matrix? That is, how do we know we are not simulated beings existing in a simulated world created with technologies vastly in advance of our own, by intelligent beings—whether human, extraterrestrial, biological, or artificial? The answer, of course, is that we don’t know, and it may be that we can’t know. At least not easily. Of course, the simulators might deliberately reveal themselves, or they might get careless. (“Hi. Some of you may be wondering why the sky turned to static yesterday. This is kind of embarrassing, but . . .”) But there might be another way. Even simulators very careful to hide their tracks might, from time to time, leave traces.
Suppose the simulators wanted to make it impossible for inhabitants of a simulated universe to ascertain with certainty that their universe was simulated. The simulated universe would not need to be infinite; it would need to extend outward from the inhabitants only as far as their instruments could probe. In other words, it would need to be only the size of an observable universe. If you’re wondering what it might feel like to be a member of that suitably advanced civilization, try saying “only the size of the observable universe” a few times in a pleading what-I’d-really-like-for-my-birthday tone. More seriously, we may be getting cavalier with our assumptions. It’s possible that even for the civilization we are imagining, an observable universe-sized simulation would be an ambitious undertaking. But it could be made less ambitious.
If the simulators wished to save on costs (or scrimp on birthday presents), they would realize they needn’t simulate all the universe’s parts, but only the parts its inhabitants happened to be looking at or otherwise experiencing at any given moment. This would mean that simulated matter on the molecular scale could be left unresolved except in places where an inhabitant was using an electron microscope, and stars and galaxies in deep space could also be left unresolved except where an inhabitant happened to be pointing an optical, infrared, or radio telescope. Of course, real molecules and distant stars and galaxies would in various ways affect the inhabitants’ nearer environs, and the inhabitants with some knowledge of molecular theory and gravity would know how to detect and measure those effects. But the simulators could approximate them, and the inhabitants would be none the wiser.
Approximations, though, come at a cost. Over time they would grow into inconsistencies large enough to crash the program. To prevent a crash, the simulation would require occasional patching. Things would never be perfect. Even with patching, if the inhabitants looked hard enough they might detect inconsistencies, small tears in the cosmic scenery. Exactly what would those tears look like? John Barrow thinks they might look like that (possible) small change, 6 billion years ago, in the strength of the electromagnetic force14—a prospect that may raise the possibly simulated hairs on the back of your possibly simulated neck.
To most of us, the idea that we are simulations is fairly abhorrent, but that is not a reason to think it unreal. In fact, the Oxford philosopher Nick Bostrom makes a persuasive case. Bostrom reasons that members of a society with a sufficiently advanced technology, a society that can create simulated universes, will create a great many. Sooner or later inhabitants of simulated universes will outnumber inhabitants of actual universes, and their population will only increase over time. Enter, once again, the principle of mediocrity. You can probably see where this is heading. If we do not know whether our universe is real or simulated but we have reason to believe that simulated universes vastly outnumber real ones (suppose, for instance, there are a thousand simulated universes for every real one), then we’ve no choice but to conclude that odds are we are living—er, “living”—in a simulated universe.
We can’t know the sorts of universes or inhabitants that society with the suitably advanced technology prefers to create, but we might with some reason expect that its members wish to be edified and entertained. It’s reasonable to assume that if they can begin simulations, they can also end them—and they may do so for many reasons, including waning interest. So, if that possible change in the electromagnetic force 6 billion years ago has you worried, you might take it as incentive to keep the simulators interested. Exactly what might the simulators find interesting? Obviously we can’t know, but if they are at all like the players of computer games I’ve seen, they bore easily, they seldom reward good behavior, and they almost never reward passivity. If we’d like our universe to continue running, perhaps we all should start taking more risks.
Readers who haven’t put down the book to make a reservation to bungee jump or take up extreme downhill skiing will be comforted to know that simulators might program the simulation for risk in many ways, most of which pose no threat to our well-being. One of these brings us back to weird life. Let’s back up on the speculative branch a bit, and assume that we inhabit a real universe. But let’s also assume that as a more or less inevitable by-product of technological advances, computer simulations will continue to become ever more realistic. Simulations invite risk precisely because there is nothing real at risk. All simulators are inclined to experiment, to explore extremes, to push boundaries. This is true for a game player breaking the world land speed record on a twisting stretch of autobahn. It is also true for an entomologist manipulating virtual populations of monarch butterflies.
Might the same inclination toward risk be manifest in the organisms themselves? Perhaps. It may not be too much to expect that simulators interested in biology will have a preference for subroutines that generate simulated weird life. Even now, biologists simulate organisms and their parts; the processes of protein folding and binding, for instance, are routinely simulated with parallel and distributed computing.15 Simulated organisms in computer games, a realm mostly removed from scientific inquiry, are commonplace, and a quick census would likely show that the weird (meaning dragons and the like) would greatly outnumber the familiar.*
However, merely observing simulations and making an adjustment now and then is likely to prove a dull pastime. Sooner or later, simulators would wish to engage their simulations directly. This wish, in a manner that no doubt will one day seem primitive, is being answered even now. Will Wright, the designer of the video game simulations SimCity, The Sims, and Spore, is designing a game he calls HiveMind. It is a set of cross-platform, online applications desig
ned to turn a gamer’s everyday life into an interactive experience by tapping into personal information on phones, tablets, social networks, and computers. It will offer an experience, so Wright claims, that will merge the virtual with the real.16
Whether this is the first iteration of such a merging or a false start, it seems clear that the boundary between the real and the virtual is likely to grow ever more porous. Assuming no technical impediments, at some time in the future we’ll see two-way traffic between real and simulated environments. From the previous we might draw an interesting conclusion. If the impulse to simulate weird life is significant, and if computing power continues to increase to the degree that real persons can visit simulations and simulations can visit the real, then it follows that we or our descendants will encounter weird life directly—a prospect that yields to imaginings of griffins leaving beak marks on our ankles and unicorns following us home.
A BIT OF PERSPECTIVE
We’ve traveled a long way from speculation about bacteria with arsenic in their DNA and desert varnish as living. It may be time to catch our breaths and consider the lines of thought that brought us here. The attentive reader may have noticed that ideas for the weirdest sorts of weird life did not originate with biologists or even, for that matter, with astrobiologists. They came from scholars and practitioners in other fields. The hypotheses of life in other universes were formulated by theoretical physicists (Harnik, Kribs, and Perez; and Jaffe, Jenkins, and Kimchi). Ideas of life in the vicinity of black holes and the atmospheres of white dwarf stars were conceived by astrophysicists (Adams and Laughlin). Hypotheses of life surviving through eternity were developed by a mathematician and theoretical physicist (Dyson), who also supplied us with what may be the broadest definition of life so far. Of the many ideas of weird organisms from science fiction, two that are notably well grounded in science are from a physicist turned aerospace engineer (Forward) and a professional astronomer (Hoyle). Even the relatively conservative hypothesis of hydrogen-breathing dirigibles was proposed by a physicist (Saltpeter) and a planetary scientist turned astrobiologist (Sagan).
Many biologists who have considered alternative forms of life are inclined to agree with Norman Pace, who, you’ll recall, suspects that if life exists elsewhere in the universe, it has a biochemistry much like that of life we know. Certain astrobiologists—Schulze-Makuch and McKay, for instance—do have hypotheses for organisms that use another biochemistry; but by comparison with ideas of life on crusts of neutron stars and of life in the vicinity of black holes, they seem fairly tame. Clearly, there are two very different levels of speculation here, and between them a wide gap that needs explaining.
It’s reasonable to suppose that the reason for some of the gap is the nature of the fields of study (physics and astronomy on the one hand, and biology on the other), the perspectives that come with study in those fields, and the sort of intellect those fields attract to begin with. On the one hand, and to generalize, theoretical physicists are likely to be less interested in the particulars of a given phenomenon than they are in the underlying principles those particulars represent. Astronomers are necessarily cognizant of billions of stars as potential suns, the billions of planets now estimated to be orbiting them, and the enormous timescales over which the universe has existed and is likely to continue to exist. To a biologist pessimistic about the likelihood of weird life, a theoretical physicist might say that the particulars of biochemistry and chemistry are not as important as life’s most basic needs: energy and matter. And an astronomer would insist that those needs are met in many places in our universe (and if they exist, others), and have been met for billions of years.
On the other hand—and to generalize further—a biologist is more likely than is a theoretical physicist to be aware of the fantastically complex chemical reactions that occur within a living cell, and the improbably long series of steps, many still not fully understood, that led from amino acids to that cell. To a physicist or astronomer optimistic about the probability or inevitability of weird life, a biologist might counter that he fails to appreciate the intricacies of chemistry and biochemistry that would be necessary to produce it.*
The gap is a reminder that scientists have taken the concept of weird life seriously only recently. If it may be said to represent a new field of study, it is a field that, as was once said of astrobiology, has no subject. Or no subject yet.
* * *
* The terms “Copernican principle,” “cosmological principle,” and “principle of mediocrity” have often been used interchangeably, although the last, as defined by physicist Alexander Vilenkin, has a specific valence. It asserts that since we are typical of intelligent beings in the multiverse, our observations should be typical of those made by all such intelligent beings. (Greene, Hidden Reality, 180) Since Vilenkin’s definition most suits our purposes, I will use “principle of mediocrity” throughout.
† The world of quantum physics has aspects that are counterintuitive, one of which is its intrinsic uncertainty. A subatomic particle like an electron, for instance, cannot have a position and speed simultaneously; it is for this reason that a quantum physicist speaks not of a particle’s position and speed but of its “quantum state.” Likewise, an arrangement of particles within a volume would have an observably distinct quantum state. For our purposes the word “arrangement” will suffice.
* Alas, a common misunderstanding of scientific practice and terminology still necessitates a footnote. The oft-heard dismissal “it’s just a theory” fails to recognize the full meaning of the word “theory.” A real scientific theory is a self-consistent set of hypotheses that make predictions about nature that are testable and falsifiable. When the prediction of a certain hypothesis is shown to be correct (as when, for instance, a hypothesis of Einstein’s theory of general relativity predicted that sufficiently precise measurements would show that the Sun’s gravity was bending starlight), the theory gains credibility and authority. As more and more of a theory’s hypotheses are shown to be true, the theory is judged successful. But the theory in its entirety is not proved, and probably never will be. A theory is always provisional simply because we most likely will never know everything, and some things we do not know may someday disprove a certain hypothesis, thus collapsing the theory it helped to support and sending the hypothesizer back to the proverbial drawing board or proverbial cocktail napkin. A theory’s provisional nature, far from being a flaw, is essential to scientific advancement.
* In 1904, British naturalist Alfred Russel Wallace observed, “Such a vast and complex universe as that which we know exists around us, may have been absolutely required . . . in order to produce a world that should be precisely adapted in every detail for the orderly development of life culminating in man.” (Wallace, Man’s Place, 256–7)
* Accordingly, the Standard Model is sometimes termed the “theory of almost everything.”
* There is another explanation for the values of the constants. It also depends on odds, but odds of a different sort. In 1995, cosmologist Edward Harrison, then at the University of Massachusetts, speculated that our universe is artificial, created by an intelligence superior to ours and existing in a “mother” universe whose physical constants are similar to our own. His thinking rested on the supposition that suitably advanced civilizations, driven by a creative impulse, will wish to produce child universes and will have the means to do so. From this it follows, he wrote, that universes unfit for life cannot produce the sufficiently advanced civilizations necessary to spawn child universes. But universes fit for life can produce such civilizations, and these, in turn, may produce child universes. Since universes whose constants do not support life are not reproduced, there will be relatively few of them. Since universes whose constants do support life multiply and multiply again, there will be a great number of them, and as the multiverse grows older that number will grow, with fertile universes coming to greatly outnumber sterile universes. So, since most universes are conducive to life,
it follows that we are not living in a privileged place, but just another universe—a typical one. Furthermore, if at most times in the history of the multiverse there are a great number of universes, then we are not living at a privileged time either. Harrison had taken the problem presented by the physical constants and turned it upside down. And perhaps without intending to, he had also supplied a reason to consider our universe, which he hypothesized might be a product of random mutation and natural selection, as living. (Harrison, Cosmology)
* The same idea is postulated in Jorge Luis Borges’s 1941 short story “The Garden of Forking Paths.” A character “believed in an infinite series of times, in a dizzily growing, ever spreading network of diverging, converging and parallel times. This web of time—the strands of which approach one another, bifurcate, intersect or ignore each other through the centuries—embraces every possibility.” (Ficciones, 100)