Header 1

Our future, our universe, and other weighty topics

Wednesday, June 8, 2016

Flying Pink Unicorns and Simulated Universes

Elon Musk is very good at technology, but he is much weaker at the game of philosophical arguments, judging from his recent reasoning on the subject of whether we live in a simulated universe. Here was the argument Musk gave:
The strongest argument for us being in a simulation probably is the following. Forty years ago we had pong. Like, two rectangles and a dot. That was what games were.
Now, 40 years later, we have photorealistic, 3D simulations with millions of people playing simultaneously, and it's getting better every year. Soon we'll have virtual reality, augmented reality.
If you assume any rate of improvement at all, then the games will become indistinguishable from reality, even if that rate of advancement drops by a thousand from what it is now. Then you just say, okay, let's imagine it's 10,000 years in the future, which is nothing on the evolutionary scale.
So given that we're clearly on a trajectory to have games that are indistinguishable from reality, and those games could be played on any set-top box or on a PC or whatever, and there would probably be billions of such computers or set-top boxes, it would seem to follow that the odds that we're in base reality is one in billions.
Tell me what's wrong with that argument. Is there a flaw in that argument?
Sure, Elon, I'll tell you what the flaw is in your argument. What Musk describes is a progression of sophistication in video game technology. But we have not one bit of evidence that any computer or video game has ever itself had the slightest iota of experience, consciousness, or life-flow. We only have evidence that biological creatures such as us can have some experience, consciousness, or life-flow. So there is no basis for thinking that some super-advanced alien civilization could ever be able to produce computers or video games that by themselves were the source of experiences like the ones we have. Making such an assumption based on an extrapolation of technical progress in video games is like arguing that one day video game characters will be so realistic that they will step out of the video game screen and help us clean our house.

Biological entities such as ourselves may have experiences (or simulated experiences) when we view the output of computers or video games, but it is not the computers or video games that ever have such experiences by themselves. We have no good reason for concluding that computers or video games will ever be able to have experiences or life-flow like that we experience, so there is no reason for thinking that our experiences or life-flow is caused by such electronic devices. As for the “one in billions” part of the argument, it is just an arbitrary number that Musk has picked out of a hat, without suggesting any basis for such reasoning.

The lives of humans have produced an ocean of life-flow or experience or consciousness, but we have not the slightest evidence that any computer has produced the tiniest bit of any such thing – not even a drop, we may say. Nor is there any reason for supposing that future advances will produce such a thing. There will never be a way to program a computer so that it “has a day” or “lives an experience.”

A video game is not something that produces experiences, a fact that we can prove by imagining a world in which there are no humans, and asking: how many video game experiences would we then have? A video game is something that can affect the experiences of agents (humans) that are capable of having experiences. Arguing that a video game or computer will be able to produce experiences is like thinking (based on the fact that thunderstorms can affect baseball games) that thunderstorms will one day be able all by themselves to produce baseball games when humans aren't around.

In the original paper discussing the simulation argument that you are (or may be) living in a computer simulation, Nick Bostrom defined an “ancestor simulation” as something that would be produced when a computer would “simulate the entire mental history of humankind.” Here is the unjustified claim given in the conclusion of Nick Bostrom's original paper presenting his argument regarding the idea that you are living in a computer simulation.

A technologically mature “posthuman” civilization would have enormous computing power. Based on this empirical fact, the simulation argument shows that at least one of the following propositions is true: (1) The fraction of human-level civilizations that reach a posthuman stage is very close to zero; (2) The fraction of posthuman civilizations that are interested in running ancestor-simulations is very close to zero; (3) The fraction of all people with our kind of experiences that are living in a simulation is very close to one.

Bostrom's paper justified no such assertion. He wants us to choose between one of three possibilities, when there is a very plausible fourth possibility: The fraction of posthuman civilizations that are capable of running ancestor-simulations is zero. Since we have no evidence that computers or video games will ever be capable of producing human-like experience, this is a very plausible possibility that short-circuits his whole line of reasoning. Also, Bostrom does nothing to justify any statements about fractions or ratios, so he has just "picked out of a hat" his third possibility. 

Bostrom also makes the big mistake of implying that if there is one alien civilization interested in creating an “ancestor simulation,” that such a civilization would now be producing countless such simulations. He suggests that if there is one such civilization, the number of simulated lives would greatly outnumber the number of real lives. This is a completely unjustified insinuation. The more often some weird non-essential project has been done, the less people tend to be interested in doing it. While an alien civilization might run some ancestor simulation of the type Bostrom imagines, we have every reason to suspect that it would grow bored with such a thing after some particular number of years, and lose interest in it. Given an alien civilization that had one point in its long existence had an interest in running an ancestor simulation, there is no reason to think (given a very long life-time for that civilization) that it would now be running such simulations. And there is also no reason to believe that it would now be running very many such simulations, so many that the number of simulated lives would outnumber the number of real lives. 

I may note that Bostrom's paper is filled with quite a few goofy misstatements, such as the completely incorrect claim that "there are certainly many humans who would like to run ancestor-simulations if they could afford to do so."  Not correct, as no one prior to Bostrom's paper expressed any interest in creating computer simulations to recreate "the entire mental history of mankind." 

Reasoning like Bostrom's could be used to establish a million and one goofy claims. For example, imagine you want to establish that there is a city populated by flying pink unicorns. You could argue that super-advanced alien civilizations would have the power to create flying pink unicorns, using either genetic engineering or advanced robotics. You could then argue that since there might be trillions of super-advanced alien civilizations, probably at least one of them has had an interest in creating a city filled with flying pink unicorns. Presto! You now have your magic city filled with flying pink unicorns, located somewhere out there in the depths of space. Of course, this is the same kind of reasoning used to establish some suspicion that we are living in an alien civilization's computer simulation. It is reasoning that leaps from “they might possibly have done this weird thing” to “they probably did do this weird thing.”

There is a general problem with all such lines of reasoning. As a civilization's technology increases, the number of possibilities and opportunities available to that civilization exponentially increases. So if we imagine some super-advanced civilization with a technology millions of times greater than ours, we must imagine that the number of possibilities and opportunities open to that civilization would also increase by millions of times, or probably even billions or trillions of times. So with all those nearly infinite possibilities available to such a civilization, millions of which we might be able to imagine and billions of which we could never imagine, what is the chance that a super-advanced civilization would undertake one particular non-essential technological project we might imagine – whether it be creating “ancestor simulations” or creating cities filled with flying pink unicorns? No greater than one in a million.

So let's summarize the case against the idea that you are living in a simulated universe.
  1. There are very strong reasons for believing that experience and consciousness and life-flow such as we have cannot be electronically produced, completely ruling out the possibility that human experience is computer-generated.
  2. Even if it were possible for human experience to be electronically simulated, there is no reason to believe that any extraterrestrial civilization currently has a very strong (and abundantly funded) interest in creating “ancestor simulations” that electronically reproduce lives such as ours, given the nearly infinite number of other projects such extraterrestrials might be preoccupied with.
  3. Even if an extraterrestrial civilization did have such an interest in creating “ancestor simulations,” there is no reason to think that they would create so many such simulations that the number of simulated lives would outnumber the number of real lives.
  4. There is therefore no probability basis for assuming some substantial chance that you are part of some “ancestor simulation” created by extraterrestrials, based on a comparison between the number of simulated lives and the number of real lives. 

     This is not your life