Can you see the glitch in the numbers?
Have you ever felt like Neo in The Matrix? Like you are the only one seeing through everyone else's reality to the truth that lies behind their perceptions? Nah, me neither. ;-) But still, let's look at the perceptions involved with this week's thought experiment and see if we can spot something interesting in them.
---------------------------------------------------
Ever since the accident, Brian had lived in a vat. His body was crushed, but quick work by the surgeons had managed to salvage his brain. This procedure was now carried out whenever possible, so that the brain could be put back into a body once a suitable donor had been found.
But because fewer brains than bodies terminally fail, the waiting list for new bodies had got intolerably long. To destroy the brains, however, was deemed ethically unacceptable. The solution came in the form of a remarkable supercomputer from China, Mai Trikks. Through electrodes attached to the brain, the computer could feed the brain stimuli which gave it the illusion that it was in a living body, inhabiting the real world.
In Brian's case, that meant he woke up one day in a hospital bed to be told about the accident and the successful body transplant. He then went on to live a normal life. All the time, however, he was really no more than his old brain, kept alive in a vat, wired to a computer. Brian had no more or less reason to think he was living in the real world than you or I. How could he—or we—ever know differently?
Sources: The first meditation from Meditations by Rene Descartes, 1641; chapter 1 of Reason, Truth, and History by Hilary Putnam, 1982; The Matrix, directed by Larry and Andy Wachowski, 1999; Nick Bostrum's Simulation argument, www.simulation-argument.com.
Baggini, J., The Pig That Wants to Be Eaten, 2005, p. 151.
---------------------------------------------------
This is very familiar ground by now, so let's try to cover it quickly. A few months ago in my Response to Thought Experiment 38: I Am a Brain, I considered how unsatisfying it would be to interact with the present world as just a brain in a vat. Now, however, Baggini has upped the ante by making us consider the scenario where a "computer could feed the brain stimuli which gave it the illusion that it was in a living body, inhabiting the real world." Is such a scenario possible? Perhaps. So, by the rules of philosophical thought experiments, we must consider it, and we must consider it to be a perfect simulation with no hint of betrayal about its background operations.
Fortunately, I've already discussed the problem with this idea too. Twice. Way back in my Response to Experiment 1: The Evil Demon, we considered the (seemingly very remote) possibility that all of reality is a delusion caused by some malevolent god. Later, in my Response to Thought Experiment 28: The Nightmare Scenario, we wondered how we could know if this was all a dream that only seemed real. In the first case, I said:
Sure, the evil demon is a possibility. And so is a universe run by a Judeo-Christian god, or by a host of gods on Mount Olympus, or in an advanced civilisation's computer simulation, or in an infinite number of other imagined scenarios. But we see no evidence of this. The laws of nature don't suddenly change from one day to the next at the whim of these puppet masters. If anything is up there pulling the strings...so what? Does that mean we should do anything differently? No, it does not.
The second time around, I again nearly answered all of this week's experiment when I wrote:
Dude, what if this is all, like:
a) the Matrix and we're all actually lying in alien tubs.
b) a computer simulation from a really advanced race trying to figure out evolution.
c) a universe within a universe within a universe of nested nuclear explosions for each big bang.
d) a 3D hologram in a 6D multiverse that we don't have the senses to understand.
e) a trick being played on us by God who really set everything up in one day 6,000 years ago.
f) the dream of that bird over there and as soon as he wakes up we'll all disappear.
g) etc., etc., etc.
Whoa.... Pass me another slice of pizza bro... This week's thought experiment is the kind of head-in-the-cloud nonsense that used to plague philosophy and caused Ludwig Wittgenstein to react by saying: "The difficulty in philosophy is to say no more than we know." and "What can be said at all can be said clearly, and what we cannot talk about we must pass over in silence." ... And so what exactly are we supposed to do about any entry on the infinite list of infinitely remote possibilities that we're all just really acting under a monstrously big and fiendish illusion? Should we all give up trying to live good lives because, hey, you never know if the answer to life is really, a, b, c, d, e, f, or g from above? Of course not. So let's just move on and deal with the world as we see it.
In some back-and-forth that followed this analysis, I was required to point out the difference between productive speculation, which is the hallmark of good science, and pernicious speculation that specifically eliminates the possibility of testable hypotheses. The second one is the kind that is pointless to discuss. I mentioned the computer simulation argument in each of my previous responses because I consider that to be another of these pernicious examples, but I didn't address it fully at the time. Now's my chance, however, as Baggini wants us to consider it specifically. He does a good job in his discussion of this thought experiment of summarising the original argument from Nick Bostrom, so lets start with that:
Indeed, a recent argument has even suggested that it is overwhelmingly probable that we are living in a virtual reality environment, not perhaps as brains in vats, but as artificially created intelligences. The argument is that, given time, we or another civilisation will almost certainly be able to create artificial intelligences and virtual-reality environments for them to live in. Further, because these simulated worlds do not require the huge amount of natural resources to keep them going that biological organisms do, there is almost no limit to how many such environments could be created. There could be the equivalent of an entire planet Earth 'living' in one desktop computer of the future. If all this is possible, we have only to do the maths to see that it is probable we are in one such simulation. Let us say that over the whole course of human history, for every human being that ever lives, there are another nine that are the creation of computer simulations. Both the simulations and the humans would believe that they are biological organisms. But 90 per cent of them would be wrong. And since we cannot know if we are simulations or real beings, there is a 90 percent chance that we are wrong to think we are the latter. The question we need to ask is not whether it sounds incredible, but whether there is anything wrong with its logic. And identifying its flaws is a very difficult, if not impossible, task.
Impossible?? Pshaw. In some of the nineteen academic papers listed on Bostrom's site that discuss the simulation argument—(nineteen!?!?)—none made any headway with this, but a few scientists debated whether such a simulation is actually possible and discussed what flaws we might expect to see according to our understanding of physics. That is reasonable, if highly speculative, speculation. But, as I showed by adding the bolded emphasis above, this is beside the point. The simulation is indeed hypothesised to be so perfect as to be untestable. So now what? Luckily, we don't need testing to show the simulation argument is still flawed and therefore pointless.
As Baggini says, "we have only to do the maths to see that it is probable we are in one such simulation." But, like most English humanities students, he probably stopped taking mathematics courses around the age of 13. Otherwise he (or someone!) would have seen through this already. The probabilities would only show we were "likely" to be in a simulation if the only possible choices for reality were "Computer Simulated Reality" or "The One Actual Reality". But as we've seen in other thought experiments, there is an infinity of other untestable possibilities too. The mathematical argument that Bostrom makes completely breaks down under this dueling list of infinities. Therefore, there is no overwhelming possibility for any of these alternative explanations for our existence. We are still left with the parsimony of Occam's razor, and the evidence in front of us, that reality is best considered simply as it appears.
What do you think? Did you see the math error too? If so, then you too can be like Neo and be "the one" in this Mai Trikks. Sit back and enjoy this 2-minute climactic scene of him stopping bullets and imagine yourself waiting for more flawed philosophical arguments to come flying at you in the weeks ahead...
---------------------------------------------------
Ever since the accident, Brian had lived in a vat. His body was crushed, but quick work by the surgeons had managed to salvage his brain. This procedure was now carried out whenever possible, so that the brain could be put back into a body once a suitable donor had been found.
But because fewer brains than bodies terminally fail, the waiting list for new bodies had got intolerably long. To destroy the brains, however, was deemed ethically unacceptable. The solution came in the form of a remarkable supercomputer from China, Mai Trikks. Through electrodes attached to the brain, the computer could feed the brain stimuli which gave it the illusion that it was in a living body, inhabiting the real world.
In Brian's case, that meant he woke up one day in a hospital bed to be told about the accident and the successful body transplant. He then went on to live a normal life. All the time, however, he was really no more than his old brain, kept alive in a vat, wired to a computer. Brian had no more or less reason to think he was living in the real world than you or I. How could he—or we—ever know differently?
Sources: The first meditation from Meditations by Rene Descartes, 1641; chapter 1 of Reason, Truth, and History by Hilary Putnam, 1982; The Matrix, directed by Larry and Andy Wachowski, 1999; Nick Bostrum's Simulation argument, www.simulation-argument.com.
Baggini, J., The Pig That Wants to Be Eaten, 2005, p. 151.
---------------------------------------------------
This is very familiar ground by now, so let's try to cover it quickly. A few months ago in my Response to Thought Experiment 38: I Am a Brain, I considered how unsatisfying it would be to interact with the present world as just a brain in a vat. Now, however, Baggini has upped the ante by making us consider the scenario where a "computer could feed the brain stimuli which gave it the illusion that it was in a living body, inhabiting the real world." Is such a scenario possible? Perhaps. So, by the rules of philosophical thought experiments, we must consider it, and we must consider it to be a perfect simulation with no hint of betrayal about its background operations.
Fortunately, I've already discussed the problem with this idea too. Twice. Way back in my Response to Experiment 1: The Evil Demon, we considered the (seemingly very remote) possibility that all of reality is a delusion caused by some malevolent god. Later, in my Response to Thought Experiment 28: The Nightmare Scenario, we wondered how we could know if this was all a dream that only seemed real. In the first case, I said:
Sure, the evil demon is a possibility. And so is a universe run by a Judeo-Christian god, or by a host of gods on Mount Olympus, or in an advanced civilisation's computer simulation, or in an infinite number of other imagined scenarios. But we see no evidence of this. The laws of nature don't suddenly change from one day to the next at the whim of these puppet masters. If anything is up there pulling the strings...so what? Does that mean we should do anything differently? No, it does not.
The second time around, I again nearly answered all of this week's experiment when I wrote:
Dude, what if this is all, like:
a) the Matrix and we're all actually lying in alien tubs.
b) a computer simulation from a really advanced race trying to figure out evolution.
c) a universe within a universe within a universe of nested nuclear explosions for each big bang.
d) a 3D hologram in a 6D multiverse that we don't have the senses to understand.
e) a trick being played on us by God who really set everything up in one day 6,000 years ago.
f) the dream of that bird over there and as soon as he wakes up we'll all disappear.
g) etc., etc., etc.
Whoa.... Pass me another slice of pizza bro... This week's thought experiment is the kind of head-in-the-cloud nonsense that used to plague philosophy and caused Ludwig Wittgenstein to react by saying: "The difficulty in philosophy is to say no more than we know." and "What can be said at all can be said clearly, and what we cannot talk about we must pass over in silence." ... And so what exactly are we supposed to do about any entry on the infinite list of infinitely remote possibilities that we're all just really acting under a monstrously big and fiendish illusion? Should we all give up trying to live good lives because, hey, you never know if the answer to life is really, a, b, c, d, e, f, or g from above? Of course not. So let's just move on and deal with the world as we see it.
In some back-and-forth that followed this analysis, I was required to point out the difference between productive speculation, which is the hallmark of good science, and pernicious speculation that specifically eliminates the possibility of testable hypotheses. The second one is the kind that is pointless to discuss. I mentioned the computer simulation argument in each of my previous responses because I consider that to be another of these pernicious examples, but I didn't address it fully at the time. Now's my chance, however, as Baggini wants us to consider it specifically. He does a good job in his discussion of this thought experiment of summarising the original argument from Nick Bostrom, so lets start with that:
Indeed, a recent argument has even suggested that it is overwhelmingly probable that we are living in a virtual reality environment, not perhaps as brains in vats, but as artificially created intelligences. The argument is that, given time, we or another civilisation will almost certainly be able to create artificial intelligences and virtual-reality environments for them to live in. Further, because these simulated worlds do not require the huge amount of natural resources to keep them going that biological organisms do, there is almost no limit to how many such environments could be created. There could be the equivalent of an entire planet Earth 'living' in one desktop computer of the future. If all this is possible, we have only to do the maths to see that it is probable we are in one such simulation. Let us say that over the whole course of human history, for every human being that ever lives, there are another nine that are the creation of computer simulations. Both the simulations and the humans would believe that they are biological organisms. But 90 per cent of them would be wrong. And since we cannot know if we are simulations or real beings, there is a 90 percent chance that we are wrong to think we are the latter. The question we need to ask is not whether it sounds incredible, but whether there is anything wrong with its logic. And identifying its flaws is a very difficult, if not impossible, task.
Impossible?? Pshaw. In some of the nineteen academic papers listed on Bostrom's site that discuss the simulation argument—(nineteen!?!?)—none made any headway with this, but a few scientists debated whether such a simulation is actually possible and discussed what flaws we might expect to see according to our understanding of physics. That is reasonable, if highly speculative, speculation. But, as I showed by adding the bolded emphasis above, this is beside the point. The simulation is indeed hypothesised to be so perfect as to be untestable. So now what? Luckily, we don't need testing to show the simulation argument is still flawed and therefore pointless.
As Baggini says, "we have only to do the maths to see that it is probable we are in one such simulation." But, like most English humanities students, he probably stopped taking mathematics courses around the age of 13. Otherwise he (or someone!) would have seen through this already. The probabilities would only show we were "likely" to be in a simulation if the only possible choices for reality were "Computer Simulated Reality" or "The One Actual Reality". But as we've seen in other thought experiments, there is an infinity of other untestable possibilities too. The mathematical argument that Bostrom makes completely breaks down under this dueling list of infinities. Therefore, there is no overwhelming possibility for any of these alternative explanations for our existence. We are still left with the parsimony of Occam's razor, and the evidence in front of us, that reality is best considered simply as it appears.
What do you think? Did you see the math error too? If so, then you too can be like Neo and be "the one" in this Mai Trikks. Sit back and enjoy this 2-minute climactic scene of him stopping bullets and imagine yourself waiting for more flawed philosophical arguments to come flying at you in the weeks ahead...
RSS Feed
