Lead Us Not Into Speculation, Nor Excessive Computation
A prominent philosopher argues that you, me, and everyone you know may be an artificial computer-simulation of a person.
More Matrix and PhilosophyPublisher: Open Court
Subtitle: Revolutions and Reloaded Decoded
Contributors: William Irwin, Editor
US publication date: 2005-02
If there had been a Best-of-the-year list for philosophy in 2007, Nick Bostrom’s so-called simulation argument would have been near the top. The European press has been all over Bostrom’s claim for a few years, but it hit big in the US on 14 August when John Tierney of the New York Times discussed it in his science column. Then Keith Olbermann picked it up on his MSNBC show. Olbermann got right to the point when he closed his segment and quipped, “I wonder what the Intelligent Design folks are going to think of this!”
This is Bostrom’s claim: that there’s a real chance -- one in five, he roughly estimates -- that me, you, and everyone else you know is a computer simulation of a person, not too unlike those characters in The Matrix. Bostrom presents his argument in the book, More Matrix and Philosophy: Revolutions and Reloaded Decoded, excerpted below, but I think there’s a shortcut to seeing both the plausibility of Bostrom’s view as well as its somewhat fragile basis.
Think of spam (the simulated email, not the simulated meat). Chances are, any particular piece of email you randomly choose in this world is likely to be spam and not a genuine letter from one person to another. That’s because spammers (apparently) can’t resist flooding the universe with their deceptive advertisements. Bostrom’s argument has a similar structure: should any civilization gain the ability to simulate conscious intelligent persons in software, he reasons, then, unless they resist the impulse, they will flood the universe with conscious, intelligent but nonetheless simulated persons. In that case, the chance is high that any person is artificial.
But is that the case? Maybe, Bostrom explains, because at least one of these claims is true: no civilization survives to create artificial life; they do survive but they resist the impulse; or we’re all probably synthetic surges of one’s and zero’s chugging along in some computer network.
So what would the Intelligent Design people think? I bet that they love the simulation argument, for the more credible and popular the idea that we live in a computer program designed intelligently by some advanced, super smart programmer, the more credible may seem their contention that evolution is guided by some superior, intelligent designer. That’s ironic, because Bostrom is not talking theology, just computer science (and some sociology and psychology) from a very long-range point of view. Yet it’s understandable that the argument points to theological questions about immortality (uploading yourself again, Mrs. Smith?) and the foundations of ethics (Sorry, the master programmer made me do it).
In fact, Bostrom’s argument, despite its empirical and scientific pedigree, indulges one idea that spiritualists, theologians, and conspiracy theorists of various kinds have long embraced (even though there is not much good evidence for it): the idea that everything we see and touch and everything that empirical science tells us is in fact an illusion, or perhaps just a tiny corner of something enormously bigger, rationally planned, and utterly different than it now seems.
Since I’m an empiricist, top honors for past year in philosophy must go to a trend that itself warns us against this temptation: the continuing rise of popular atheism evidenced by best-selling authors like Richard Dawkins and Christopher Hitchens. As David Ramsay Steele reminded me (when discussing his forthcoming Atheism Explained), atheists tend to be empiricists who shy away from leaps of faith whether they involve theology or, in Bostrom’s case, scientific speculation. For there is a fourth possibility that Bostrom acknowledges, but leaves out of his presentation: the fourth possibility that no amount of computational power or technological sophistication is sufficient to create conscious artificial life. If that’s so, the simulation argument becomes a mere simulation of a good argument -- unless of course some clever programmer thought it would be fun to have some “person” conclude just that.
Below is extracted from “Why Make a Matrix? And Why You Might Be In One”, by Nick Bostrom in More Matrix and Philosophy: Revolutions and Reloaded Decoded, edited by William Irwin and published by Open Court Publishing Company, 2005.
The Simulation Argument
If each advanced civilization created many Matrices of their own history, then most people like us, who live in a technologically more primitive age, would live inside Matrices rather than outside them. If this were the case, where would you most likely be?
The so-called Simulation argument, which I introduced a few years ago, makes this line of reasoning more precise and takes it to its logical conclusion. The conclusion is that there are three basic possibilities at least one of which is true. The first possibility is that the human species will almost certainly go extinct before becoming technologically mature. The second possibility is that almost no technologically mature civilization is interested in building Matrices. The third possibility is that we are almost certainly living in a Matrix. Why? Because if the first two possibilities are not the case, then there are more “people” living in Matrices than in “real worlds.” As a “person” then the chances are that you are living in a Matrix rather than in a “real world.”
The Simulation argument does not tell us which of these three possibilities obtain, only that at least one of them does. The argument employs some math and probability theory, but the basic idea can be understood without recourse to technical apparatus. (For the full story, see “Are You Living in a Computer Simulation?” Philosophical Quarterly, Vol. 53, No. 211 (2003), pp. 243–255. This and other related papers are available at www.simulation-argument.com.)
Given that the Architects of a technologically mature civilization could create a vast number of Matrices even by devoting just a small fraction of their resources to that end, an interesting implication follows. Consider the set of civilizations that are at similar level of technological development as our own current civilization. Suppose that some non-trivial fraction of these eventually go on to become technologically mature. Suppose, furthermore, that some non-trivial fraction of these devote a non-negligible proportion of their resources to building Matrices. Then most people like us live in Matrices rather than outside them. There are thus three basic possibilities: either almost every civilization like ours go extinct before reaching technological maturity; or almost every mature civilization lacks any interest in building Matrices; or almost all people with our kind of experiences live in Matrices.
Let us think a little about these three possibilities. If almost every civilization at our current stage goes extinct before becoming technologically mature, then our future looks relatively bleak. For if such a premature ending were the fate awaiting most civilizations, we would have to suspect that the same will hold for our civilization in particular. This is because we seem to lack any reason for thinking that our civilization will be luckier than most other civilizations at our stage.
The second possibility is less depressing. It might turn out that almost all technologically mature civilizations lose interest in building Matrices. Maybe the potential Architects of the future will not share any of the possible motives for building Matrices that we discussed above. Presumably, Architects would have used their advanced technology to improve their own capacities, so they may be superintelligent and have complete control over their own mental states. Rather than resorting to Matrix-building for recreation, they may obtain pleasure more efficiently by direct stimulation of their brains’ pleasure centers. Their science may be so advanced that they have little to learn from running simulations of their historical past. Furthermore, they might develop ethical norms that prohibit the creation of Matrices. So we cannot infer from the fact that many current people would be tempted to construct Matrices that the same would hold for the super-advanced folks that would actually have the ability to act on this motive.
The third possibility is the most intriguing. If the vast majority of all people with other kind of experiences live in Matrices then we probably live in a Matrix. Unless we had some specific evidence to the contrary, we would therefore have to conclude that the world we see around us exists only by virtue of being simulated on a powerful computer built by some technologically highly advanced Architect.
Not the Old Brain-in-a-Vat Argument
For hundreds of years, philosophers have pondered the question how we can know that the external world exists. René Descartes (1596–1650) posed this question in his Meditations, and considered the scenario where a hypothetical evil demon caused us to have erroneous beliefs about external objects. In more recent years, Descartes’s skeptical scenario has been given a more modern finish, and instead of a demon one is now asked to imagine a mad scientist who has extracted one’s brain and who keeps it in a vat where the scientist is stimulating it with electrical signals replicating the sensory input that the brain would have had if it had interacted with a very different environment from that which is present in the real world. This is, of course, is the predicament explored in the Matrix movie. How can one possibly know that one is not such a brain in a vat, the philosophical skeptic challenges, given that all the appearances we experience could be the experiences of an envatted brain?
The argument outlined above provides a much stronger reason for taking seriously the possibility that we are living in a Matrix. The traditional skeptical argument offers no positive ground for thinking that we are living in a Matrix. At best, it shows that we cannot completely rule out that possibility, but we remain free to assign it a very small or negligible probability. If there are no mad scientists who experiment on conscious envatted human brains, then we are not envatted. Even if there were a few such brains-in-vats, they might be extremely rare compared to the brains-in-crania that interact with the external world in the normal way; and if so, then it may be highly unlikely that we would be among the envatted ones.
The Simulation argument, by contrast, adopts as its starting point that things are the way they seem to be and that science gives us reliable information about the world. Part of this information concerns the technological capabilities that an advanced civilization would be able to develop. Among these would be the capability to create Matrices. Crucially, it seems that they could easily create Matrices in astronomical numbers. From this we can then conclude that either technologically mature civilizations that are interested in creating Matrices are extremely rare compared to civilizations at our own current stage of development or almost all people like us live in Matrices. And from this, the division into three the three basic possibilities mentioned above follows.
The Simulation argument itself doesn’t tell us which one of these three possibilities obtain. In fact, we do not currently have any strong evidence either for or against either of these three possibilities. We should therefore assign them all a significant probability. In particular, we should take seriously the possibility that we are living in a Matrix. We might still think that the probability is less than 50 percent. A degree of belief of something like 20 percent would seem quite reasonable given our current information.
How to Live in a Matrix
The Simulation argument does have some more subtle practical ramifications, even if we set aside the other two possibilities to which it points (which do not entail that we are in a Matrix). Some scenarios that would otherwise seem to have been foreclosed by our current scientific understanding again become real possibilities if we inhabit a Matrix. For instance, while the physical world cannot suddenly pop out of existence, a simulated reality could do so at any time if the Architect decides to pull the plug. An afterlife would also be a real possibility. When a person dies in a simulation, he or she could be resurrected in another simulation, or the Architect could uplift the deceased into his own level of reality.
It is also conceivable that only some people are simulated in enough detail to be conscious while others may be simulated at a cruder level allowing them to appear and behave much like the real people but without having any subjective experience. The so-called “problem of other minds” -- how we can know that other people are really conscious and are not just behaving as if they were -- is another old chestnut of philosophy. There is, however, no consensus that such “zombie” people are possible even in principle. Some people have argued that it is necessarily true that anybody who acts sufficiently like a normal human being must also have conscious experience. (Whether this view would entail that your least favorite politicians cannot be zombies is a question on which more research is required.)
Another possibility it that the Architect might decide to reward or punish his simulated creatures, perhaps on the basis of moral criteria. If you might be in a Matrix, this consideration may give you a novel self-interested reason for behaving morally. The situation would be analogous to the case where God is watching and judging you except that the role of the final judge would not be a supernatural being but the physical person or persons who built the Matrix.
It would be misleading to say that if we are in a Matrix then we and the world around us do not really exist. It would be more accurate to say that the reality of these things is of a somewhat different nature than we thought before. Your nose would still be real; only, its reality would consist in being simulated on a powerful computer. The computer and the electrical activity of its circuitry would be physical phenomena in the more basic level of reality inhabited by the Architect of the Matrix.
At a minimum, the Simulation argument provides many exciting avenues for philosophical thinking. But if it is sound -- and so far it has not been refuted -- it could also provide various suggestions, however tentative and ambiguous, for how we should go about our lives and for what we should expect in the future. When we follow through the logical implications of what we think we know, we discover just how much we don’t yet know.
George Reisch is the series editor for Open Court's Popular Culture and Philosophy series. He received a Ph.D. in History and Philosophy of Science from the University of Chicago in 1995 and teaches philosophy at the School for Continuing Studies at Northwestern University. His book, How the Cold War Transformed Philosophy of Science, was published by Cambridge University Press in 2005.
Nick Bostrom is a philosopher at Oxford University and Director of the Future of Humanity Institute.