Simulation hypothesis

Financial cosmology The JC’s guide to theoretical physics in the markets.™ →Go uptown Index: Click ᐅ to expand: Cosmology Lexophysics Tell me more Sign up for our newsletter — or just get in touch: for ½ a weekly 🍺 you get to consult JC. Ask about it here.

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.

If (1) is true, then we will almost certainly go extinct before reaching posthumanity. If (2) is true, then there must be a strong convergence among the courses of advanced civilizations so that virtually none contains any relatively wealthy individuals who desire to run ancestor-simulations and are free to do so. If (3) is true, then we almost certainly live in a simulation. In the dark forest of our current ignorance, it seems sensible to apportion one’s credence roughly evenly between (1), (2), and (3).

—Nick Bostrom, Are You Living in a Computer Simulation? (2003)

I wish I could summon a strong argument against it, but I can find none.

— Neil Degrasse Tyson

“I speak of none but the computer that is to come after me,” intoned Deep Thought, his voice regaining its accustomed declamatory tones. “A computer whose merest operational parameters I am not worthy to calculate—and yet I will design it for you. A computer that can calculate the Question to the Ultimate Answer, a computer of such infinite and subtle complexity that organic life itself shall form part of its operational matrix.

— Douglas Adams, The Hitch-Hiker’s Guide to the Galaxy

An amusing, but fundamentally preposterous a priori argument which purports to prove by deduction, in the same way that Rene Descartes deduced the existence of rice pudding and income tax, that either we are as good as dead, or we live in a Matrix.

Spoiler: a priori arguments are conjuring tricks. They are fun and entertaining. But don’t try them at home.

This one is practically impossible to try at home, of course — which is, perhaps, why intelligent people like Neil Degrasse-Tyson who ought to know better — fall for it.

The argument

1. If you accept philosophical materialism [i.e., that there is no God and that human consciousness is not the product of some sort of non-material “spirituality”] and you have a sufficiently powerful computer, you can emulate human consciousness.
2. If you can emulate human consciousness, you can simulate the existence of people like your forebears. In other words, you are not (necessarily) emulating people as clever as you technologically-advanced pan-dimensional hyperbeings who actually built the Matrix, but just the dopey 20th century meatsacks you evolved from: that is, people like us.
3. A sufficiently fine-grained, conscious, simulated dopey 20th century meatsack would be unable to tell itself apart from a biological dopey 20th century meatsack, and vice versa.
4. If you have enough sufficiently powerful computers, you can run a great many emulations of your dopey meatsack forebears.
5. If you ran enough emulations, simulated humans would vastly outnumber biological humans.
6. If simulated humans vastly outnumbered biological ones, and one could not tell whether one was biological or simulated, a rational human should assume itself to be simulated.
7. If you have got this far three alternative propositions is true: either
   1. any intelligent species would almost certainly wipe itself out before gaining the technology to build a matrix (the “we are the dead” option);
   2. almost no technologically mature civilisations would be bothered to even run a Matrix (the “they will be the disinterested” option); or
   3. You are almost certainly in the Matrix right now. (the “we are Neo” option)
8. Case closed: we’re in the Matrix.

Feel like you’ve had your pocket picked?

We are the dead

You needn’t travel very far to take the simulation hypothesis to its logical, absurd, conclusion.

It is taken to follow from the first premise that replicating a “carbon based” human computer outside an actual homo sapiens is an engineering problem, not a conceptual one. Therefore we will either do it, die before we solve the engineering problem, or be so clever by the time we could do it, that we have more interesting things to do with our technology.

Bostrom assumes that someone would be curious enough to try to build a Matrix filled with dopey 20th century meatsacks. This seems presumptuous: it is not obvious that dopey 20th century meatsacks would bother running experiments that would have fascinated people living in the stone age, but okay; let’s go with it.

If you conclude intelligent life is capable of creating a “Matrix” then, assuming that computation power scales infinitely and in inverse proportions to the required energy — and it does not seem to — there are likely to be countless simulations, only one “real thing”, so the odds are so close to certain that it doesn’t matter that you are a simulation, so you aren’t biologically alive either.

This is a neat trick: it is not to say that your material existence is impossible, but just that its probability is vanishingly unlikely.

If a Matrix isn’t possible, then — clearly — we can’t be in a Matrix, but we also must be incapable of developing a difference engine that could create one, so you wouldn’t be having this conversation in the first place.

Our materialist philosophy of mind tells us our conscious machines — brains — are purely physical; our laws of thermodynamics tell us no energy is gained or lost in our physical systems that would indicate spiritual intervention. There seems to be nothing about the carbon substrate that is makes it uniquely capable of generating consciousness.

A Cartesian slight of hand: We aren’t dead, yet, so we must therefore be in a Matrix.

Deep Thought successor redux

But, problem: unless intelligent life can eventually simulate itself — that is, create a Matrix — there can be no Matrix. The definition of a Matrix is that it is a simulation of intelligent life. If there is no intelligent life to simulate, then, whatever a Matrix is, it can’t be a simulation. Matrices are clever, but they can’t bootstrap themselves into existence — not if you want to stick with your materialist assumptions we made earlier. If there is a Matrix, there must be intelligent biological life to have invented it.

But wait: could the Matrix be — the real thing? Like, could the Universe itself be some kind of super computer?

We call this the “Douglas Adams objection”: once we get hip to substrate neutrality, isn’t the universe itself a giant computer? If so, what need of a simulation? Doesn’t a copy of the universe, doing what the universe is already doing by itself, fall rather foul of Occam’s razor? Have we just proved that life exists? If so, then I have some news, fellows: someone beat you to this splendid a priori idea by 400 years: Rene Descartes. Is not the simulation hypothesis another way of saying, “cogito, ergo sum”?

When the JC was a lad his Dad, Old Grumpy Contrarian, went snorkelling off a boat in the Ionian Sea. After a while the old man emerged, declaring triumphantly he had found an ancient anchor on the sea bed: a relic, no doubt of Siege of Troy, or the Odyssey. The whole Contrarian family was very excited.

“There is a warp tied to it!” Grumpy exclaimed, excitedly disappearing below the waves with another colossal splash. Bravely he followed the anchor warp for twenty yards across the sea floor, until it rose up, into the warm upper waters, broke the surface and revealed itself to be — attached to the runner over the prow of his own boat.

Real life meets the definition of “a computer simulation”, especially if you go substrate neutral

For this to work, the simulation would not just have to be very good: it would need to be identical to real human sentience, in every respect. Any shortcuts would lead to potential variances, and as we know from our modern morality tales about, butterflies and rain forests, jointed pendulums and so on, any atomic variations in initial conditions have colossal, non-linear knock-on effects.

This would involve not just a perfectly accurate — that is to say transcendently true — theory of human cognitive activity, but a perfectly accurate — that is to say transcendently true — theory of all events in the universe. These theories would not be models in any sense of the word, but actual replications of the actual world, that is to say, the territory itself, not a mere map.

So, firstly, the sheer computing power required to run this algorithm would be so great as to not only be practically impossible, but theoretically impossible. In fact, its operation would not so much skew the functioning of the real-world, but but duplicate it: but but you cannot duplicate the energy in a closed physical system without violating the laws of thermodynamics — unless the real world counts as a computer simulation. Which, on this logic, it does. If a computer simulation is indistinguishable from the universe itself, then the universe is the computer simulation and this hypothesis is — semantics.

And about those laws of thermodynamics: in order to draw a complete, functioning, comprehensive theory of the universe, one must first have comprehensive, true, knowledge of the total canon of all laws of science. But science being an inductive process, and quite incapable of establishing anything by way of proof, this is again theoretically impossible. Our present state of knowledge of the laws of the universe is contingent and incomplete.

Until the End of the Universe. The point at which all races are run, all data gathered, the universe has equalised itself into entropic warm brown sludge — at that point we have enough data to run our simulation. Or would do, had the data not degraded into entropic warm brown sludge, meaning there was no such data, nor any energy left to run the algorithm required of it.

The great irony of scientism is that, collected data being historical and all, and therefore necessarily incomplete (and for all intents and purposes, so small a proportion of the total quantity of information in the universe as to be statistically meaningless, that the point where these greedy reductionists can, on their own theory, run their experiments will never arrive. The run will long since have exploded, rather wrecking their simulation algorithms.

See also

• The Hitch-Hiker’s Guide to the Galaxy
• Substrate
• A priori
• Rene Descartes
• Data modernism
• Conway’s Game of Life