One important principle of any sensible social epistemology is that the fraction of crazy-sounding ideas that are really crazy is extremely high. Of course, a lot of crazy-sounding ideas have turned out being right (e.g., the evolution of different species from common ascent, the earth’s being a planet turning around a star, the atomic composition of matter, etc.), but for each one of these ‘victories of ingenuity against common-sense intuition’, literally thousands of foolish claims have existed and will exist. This means that you are not simply behaving as a dull reactionary when you instinctively dismiss some idea as ‘crazy’ if you clearly see it conflicts with common sense, but only that a sound bit of skepticism is working within your own brain. Extravagant claims need extraordinary proofs, and your natural skepticism is allowed to retreat only when some of these proofs start to be presented.
Unfortunately, the crowd of philosophers, in pursuing our legitimate social and intellectual function of testing and pushing out the limits of common sense, have very often misspelled the recommendation I have put in italics, and have interpreted it as saying something like that extraordinary claims need extravagant proofs. More than a millenium serving as ancilla theologiae (theology’s servant) and producing an innumerable array of extremely clever and logically very subtle arguments about the existence and properties of god, angels, demons, saints and souls, has surely left in some of our cleverest philosophers an ineradicable tendency to take a little bit too seriously some extravagant conjectures. We have also not to forget that, in matters of fact, ‘extraordinary proofs’ can only come from empirical findings, especially from extremely implausible predictions. A clever verbal argument, no matter how sophisticate it looks, can never serve as a proof of anything that is not a platitude. Hence, the chances that a philosopher supports a crazy-sounding idea just because the idea is ‘sexy’, rather than because there are sound reasons to support it, tend probably to be higher than your highest estimates.
I beg your pardon for having started in a so skeptical mood, but I think it was a necessary caveat before attacking a topic so charged of intellectual vehemence as the one I have chosen for this entry. Of course, neither Nick Bostrom 1 nor Elon Musk have been the first ones in publicising the idea that the world we experience may be a kind of fiction. In the tradition of western philosophy, bot Plato and Descartes are famous for suggesting something like that, the former with his ‘myth of the cave’ and the latter with his ‘evil daemon’; but the idea has still a longer tradition in the East (e.g., Maya’s veil). The revival of the conjecture that we live in a kind of illusory reality owes a lot, of course, to the growing industry of computer games and virtual reality devices, as well as to its diffusion through popular culture in movies like Matrix or Total Recall. We could say that, by the beginning of the century, the world was ripe to welcome some intellectual dignification of this fashionable trend. What might be better than a logical or mathematical proof of the reasonableness of the idea? Of course, taking into account the technology-freaky character of most of the target audience of that argument, that would be much better than a nearly unintelligible babbling about the ontology of simulacrums by a pedantic continental philosopher. Nick Bostrom, by then a young and promising analytic philosopher with a strong background in logic and maths, succeeded in supplying what the world was awaiting.
Bostrom’s argument, in a nutshell, is the following. Either it is extremely unlikely that humanity (or other intelligent life forms?) evolves till it has the capacity of creating perfect world-simulations (probably because it gets extinct before that), or there is something (e.g., a cultural taboo) that impedes a so evolved humanity to run perfect world-simulations though they have the technological capacity, or both of the previous hypotheses are false and hence, at some point in the future, some technologically sophisticated civilisation decides to run an astronomical number of such simulations. It seems that the first two hypotheses can be discarded as implausible (i.e., it seems not necessary that every civilisation gets extinct before reaching that capacity, nor that any technology is systematically forbidden for ever). Hence, it is most likely that, at some point in the history of the universe, some (and probably many) civilisations will start running ‘perfect’ virtual reality simulations, perhaps with the ‘scientific’ goal of observing and experimenting how the ‘people’ living in them evolve and happen to solve their problems, or just for sheer fun. But this means that, if the number of runs they do of those virtual realities is astronomical, the chances that we are living in one of them, instead of in a ‘real’ world, are also astronomical: there will be billions of simulated people for each ‘real’ person, and hence it is extremely unlikely that you are not a mere piece of a simulated world.
Before considering in detail the steps of the argument, I invite you to reflect on a similar one. As Bertrand Russell once said, it is strictly impossible to falsify the conjecture that the world has started to exist just five minutes ago in exactly the same state it was at that precise moment. Does this entail that it is equally probable that the world started to exist five minutes ago, and that it started to exist with the Big Bang, roughly 13.5 billion years ago? Perhaps you are tempted to answer ‘no’, but imagine that, instead of just considering those two options, we produce an astronomical array of conjectures: that the world started to exist five minutes and one nanosecond ago, or five minutes and two nanoseconds, or five minutes and three nanoseconds, and so on, and so forth. There is an astronomical number of possible moments in which the world might have started to exist ‘just as it was by then’, and hence, it looks as if the conjecture that it started to exist at the Big Bang has an astronomically small chance of being true. Our intelligence revolves (with good reasons) against this conclusion, just because the huge magnitude of the number of silly conjectures we have artificially produced does not make each one of them a bit less silly of what it was when there was only one silly conjecture under consideration; and this magnitude also does not help to make the combination of such innumerably silly conjectures a less silly idea, either. We just think that it is extremely much more likely that the universe started to exist with the Big Bang, than that it started to exist at any posterior moment ‘exactly like it was by then’, and our main reason is, very probably, that the laws of physics wouldn’t make any sense at all if the converse were true. I hope this argument serves to demystify a little bit the magic of big numbers on which Bostrom’s argument conjures its charm.
In the next entry I will offer more detailed counter-arguments, directly address to the content of Bostrom’s premises.
- Bostrom, N., 2003, “Are you living in a computer simulation?”, Philosophical Quarterly, Vol. 53, No. 211, pp. 243-255. ↩