Since the time of René Descartes, every cultural epoch has seen proposals of skeptical scenarios. The famous French philosopher had us entertain the diabolism hypothesis: what if everything we believe is the result of demonically inspired hallucination? (Descartes’s aim was not literal but methodological; he was trying to work his way down to some indubitable foundation for all knowledge.)
However, as mechanistic and materialist philosophies began taking hold, people were asked to consider the possibility that each of us is a brain in a vat where our thoughts, sensations, and everything else is the product of some scientist (presumably mad) probing our thinking organ. These days, having only gone further down the reductionist rabbit hole, the simulation hypothesis—the idea that everything we experience is the result of a simulation being run on the hard drives of computers—has been gaining adherents even beyond the asylum.
New Atheist Sam Harris and astrophysicist Neil deGrasse Tyson find it hard to argue against the idea that we’re not all the creation of a kid in his parents’ basement programming up our world. In an interview that can be found on YouTube, Larry King asks Tyson what difference it makes. Tyson replies it doesn’t make any difference at all.
Really? Imagine ignoring such dramatic worldview consequences.
Suppose that child programmer gets bored, as Tyson muses is possible, if not probable. Then, say, in his haphazard escapades, he decides to program your torture, just for the humor of it. Of course, there no way to rule out that possibility—no way even to say it is improbable—and hence, no way to have any confidence that at any moment you won’t be transported into some hellish underworld to have your legs boiled in a vat of molten sulfur. Or just be blipped out of existence entirely, or any other preposterous scenario you choose.
The real problem isn’t that Tyson takes the simulation hypothesis seriously but that he doesn’t take it seriously enough. If he really believes in such a possibility, he should hardly be able to sleep at night. After all, who could live believing that scenario is plausible: that nothing is real (in the way it seems), including those we love, and our experience is entirely contingent upon the whims of some finite, hormonally unstable intellect.
The truth, of course, as we all know, is Tyson does not really believe the simulation hypothesis is true or even plausible, despite what he says, and even if he cannot articulate why. We’ll return to this point.
But let us begin by noting an irony of the simulation hypothesis: atheists and religious skeptics are the ones who frequently if not exclusively entertain it—that is, by people who deny that a transcendent entity could have created the universe. But once the conversation is no longer theological, they suddenly entertain the possibility that our universe could be created by a computer programmer. Does that seem inconsistent to you? It does to me.
A minor quibble, but worth noting. Now, onto more serious concerns.
Not an argument against God’s existence
First, it should be acknowledged that the simulation hypothesis isn’t typically used as an argument against God’s existence—or if it’s presented as such, it isn’t a good one. Of course, you will occasionally hear someone say in response to another’s belief in God, “That’s all well and good, but who’s to say we aren’t just living in a simulation?”
To this you can give a simple response: even if the simulation hypothesis were possible (which, as we’ll see, it isn’t) and not absurd (which, as we’ll see, it is), it couldn’t be simulations all the way down. However, in many simulations we posit, the scenarios are physical, finite, and contingent. Hence, to explain why there are any simulations and not nothing instead, we still need that one necessary reality awaiting us at the metaphysical foundation of things: that necessary reality Aquinas argued can only be God. So, as an objection to the existence of God, the simulation hypothesis wouldn’t undermine any of the essential premises in traditional theistic arguments. In that regard, it’s irrelevant.
But once God’s existence is admitted, we can simply appeal to parsimony, a principle that says the best explanation of a thing is the one with the fewest possible assumptions, to say (very probably) we are not living in any computer simulation.
As philosopher James Anderson explains,
We’re actually faced with the choice between (hypothesis 1) there is a personal creator God and we have experiences of the real world, and (hypothesis 2) there is a personal creator God, he permitted the construction of a massive computer simulation, we are part of that computer simulation, but we don’t have experiences of the real world. It’s safe to say that H1 is far more coherent and parsimonious than H2! In what possible world—indeed, in what virtual world—would it be more rational to believe H2 than H1? (“Are We Living in a Computer Simulation?”, online at proginosko.com).
That said, most people presenting the simulation hypothesis aren’t doing so to block arguments for the existence of God. Most of them are naturalists and assume God doesn’t exist; thus, they are entertaining the hypothesis as a possibility and perhaps even probability. For that reason, we must say more.
Why we’re not living in a simulation
Let’s consider why it isn’t difficult to argue against the simulation hypothesis. We’ll consider two lines response. The simulation hypothesis can be rejected:
(1) For the same reason any other radically skeptical scenario can be rejected: the hypothesis is self-defeating because it would be irrational to believe.
(2) Because the underlying philosophy of mind falsely assumes that a computer could simulate consciousness in the first place.
The principal reason for rejecting the simulation hypothesis is the same for rejecting any radically skeptical scenario: our intuition tells us it’s absurd. The philosopher Thomas Reid talked about common sense, the sense that affirms to us, by deep intuition, the reliability of our senses and the falsity of hypotheses such as Descartes’ demon. For example, when I first told my wife about the simulation hypothesis, she responded, “That seems really absurd.” According to Reid, that is her common sense functioning properly.
All reasoning depends upon intuition: logical, mathematical, moral, etc. By intuition we mean nothing other than a deep “seeming” that something is thus and so. That it just seems like 2 + 2 = 4 (and so we believe 2 + 2 = 4) and that it just seems like my knee hurts (and so I believe my knee hurts), it just seems that the external world is real (and so I believe the external world is real and not a simulation).
We trust these intuitions when using reason to draw conclusions. If we cannot trust intuition, we cannot trust reason—really, we cannot reason at all. Fortunately, for most people, these intuitions provide such deep impressions of accuracy and truthfulness that most non-philosophers never think to question them.
However, the moment we entertain skeptical scenarios (including simulation hypotheses), reason has now called into question the reliability of intuition—namely, that our cognitive powers have been put into us by someone with intent to systemically deceive. When we doubt the existence of the external world, we’re making a philosophical argument for the claim that our cognitive powers are unreliable.
And the simulation hypothesis gives us reason to believe that our cognitive powers are unreliable. But reason is one of those cognitive faculties. The simulation hypothesis therefore gives us reason to distrust the deliverances of reason—including the simulation hypothesis. Therefore, the simulation hypothesis it is self-defeating, making it fundamentally irrational to believe. As twentieth-century philosopher Ludwig Wittgenstein said, “The moment you begin to doubt everything, you won’t get so far as doubting anything” (On Certainty, part 6, 141).
What all this means is epistemology, or theories of knowledge, must start from a position opposite Descartes: that is, by taking our cognitive powers as generally reliable. Descartes tried the opposite approach, but that just leaves us stuck in self-defeating skeptical scenarios. And the most we can do from there, as Aristotle said, is wiggle our finger.
Consciousness Is Not Reducible to Physics
The previous consideration about reason and intuition was an answer to all globally skeptical scenarios. However, let us now consider additional reasons why the simulation hypothesis is a non-starter. The simulation hypothesis depends upon certain assumptions—namely, a physicalist worldview plus a functionalist and computational theory of mind. If we have reason to believe such assumptions are false, we have reason to believe the simulation hypothesis is not just implausible but impossible.
One common assumption behind the simulation hypothesis is materialism, which maintains all of reality—including conscious reality—is the cobbling together of little physical bits (whatever those are) that lack intentionality, consciousness, teleology (directedness), and so on. Another assumption is the philosophy of mind known as functionalism, which states the mind just is what the brain just does. And finally, the assumption that what the brain does is compute, which is known as computationalism. (To be clear, computationalism is a form of functionalism, though not everyone who holds to functionalism accepts computationalism.) All these of these assumptions— materialism, functionalism, and computationalism—are demonstrably false.
While this article cannot extensively detail arguments against each of these assumptions, a brief sketch of objections should be enough to encourage confidence against the simulation hypothesis.
First, there is no reason to think consciousness can emerge from purely physical processes, no matter how complex. Our consciousness experiences are categorically different than physical properties as understood by materialists. Our mental life is suffused with intentionality and subjectivity, which comprise everything physical properties (according to physicalists) lack.
Thus, to imagine one could build a conscious, intentional experience from material that is unconscious and unintentional is as ludicrous as imagining one could construct a purple tower from a collection of entirely white blocks. Reason tells us it doesn’t matter how many white blocks a person has nor how much time they’re given to arrange those blocks into complex configurations.
Replacing white blocks with physical properties and the purple tower with consciousness, to understand this point is to see it is impossible for any entirely physical machine (such as a computer) to become conscious. As well, we must be careful not to fall for any verbal smokescreen the physicalists may deploy in their defense, such as calling consciousness emergent. Once the problem is properly understood, calling consciousness emergent doesn’t solve anything. It just relabels the mystery.
As philosopher Joshua Rasmussen has pointed out, the stuff of non-sense (meaning, stuff that is non-sensing) is simply the wrong material to create the stuff of sense: like trying to build a prime minister from prime numbers, it’s just not going to happen. The (quantitative) amount, shape, and speed of—call whatever the fundamental stuff physicals are committed to “dust”—is irrelevant to giving dust any (qualitative) sense or feeling.
From there, just as there is a construction problem in generating sense from nonsense, there is no less a construction problem in getting rationality from a nonrational base. To have rational thought we need thought that is based on reasons and sight (experience with reality). Dust is not like that—at least not according to physicalists. So, even if we could (per impossible) spin around enough dust to produce a thought—i.e., “there is a blue turtle”—it is impossible to see how that could ever be a rational thought, since it is based neither on reasons nor sight.
So that is one (not insignificant) issue. Given physicalism there is an impossible construction problem that prevents consciousness from being produced from any merely physical state, including computer states.
But we can go further and argue, as Edward Feser and James Ross have, that:
Everything physical is particular.
However, some thoughts are universal.
Therefore, some thoughts are not physical.
The first premise is uncontentious because every physical thing is restricted to this or that particular instantiation. There is, for example, this triangle drawn on this chalkboard and there is that triangle drawn on that chalkboard, but never do we encounter in the physical realm triangularity as such. Further, every physical thing in relation to its universal characteristic—for example, every triangle in relation to triangularity as such—is always to some degree imperfect (less than perfectly straight sides, etc.), however precise the tools are we use to draw it.
But we can entertain the concept of triangularity and not rely on any restricted mental image to do so, because such images will always be particular and therefore not universal (and to some degree imperfect). We grasp what triangularity is, and this grasping transcends the physical because everything physical—including brain states—is particular. Thus, one is not reducible to the other. Hence, at least some of our thinking cannot be entirely physical. Because our thoughts are something no physical thing can be (universal and determinate or exact in conceptual content) our thoughts cannot themselves be physical.
The conclusion of this argument—of which we have provided only a hasty overview—carries through to refute the simulation hypothesis, because whatever else computers are capable of, they are not capable of conceptual thought. But if computers cannot produce conceptual thought, they cannot produce beings such as us.