Note: This paper borrows heavily from Nick Bostrom; a link to his paper is in the works cited.
Descartes began his quest for knowledge by questioning the nature of the reality he perceived, eventually arriving at three possible conclusions (Descartes). The conclusion I will focus on in this paper is his dream argument, or the idea that we’re living in some kind of simulation. According to Nick Bostrom, “forecasts by serious technologists and futurologists predict that enormous amounts of computing power will be available in the future” (Bostrom) and as a consequence of this vast computing power humans will “run detailed simulations of their forebears or of people like their forebears” (Bostrom). In this paper I will demonstrate how it is difficult to prove humanity does not currently inhabit such a simulation, as we are now in the process of developing our own simulated realities which we will populate with artificial reflections of ourselves.
Death is the end of life in a physical reality, but in a virtual world death does not have the same finality. To illustrate, as the virtual worlds of the future become more and more complex the virtual actors inhabiting the simulation will need to keep pace with these progressive advancements, becoming more complex over time. Furthermore, with major enhancements in AI and a better understanding of the nature and workings of the human mind, what’s to stop us from making true-to-life virtual copies of ourselves that continue on once our physical bodies fail? The creation of vast virtual worlds filled with virtual intelligences is not a new concept, many multiplayer online games already contain simulated humans. And as we are better able to understand the workings of the human mind we will eventually become capable of building simulated intelligences indistinguishable from their biological counterparts. Similarly, as the complexity of our virtual worlds grow and our ability to create virtual copies of ourselves becomes possible, it follows that we would want to build out a virtual world for these copies to inhabit.
Conversely, even after taking these advancements into account, it seems that there is no good reason for humanity to simulate complex virtual versions of our ancestors, and there is no logical reason for humanity to build out a complex virtual universe with the sole purpose of simulating our ancestors. Even if we have an unlimited amount of computer power, why would we build and maintain a simulation dedicated to our forbears? It seems like a vain, pointless exercise.
On the contrary, people are by their very nature, vain. Moreover, humans have always been interested in the study of our ancestors, and we already run complex computer models to explore social, genealogical, and sociological questions. If we have the power to store or recreate the personalities of all those who came before us: our mothers, fathers, grandparents, and so on, why wouldn’t we? With the power to simulate and explore our history in a system that exactly reproduces the reality of our lives we will assuredly make use of it. To demonstrate, we didn’t build computers to play computer games, we built them to do complex math. Respectively, our initial virtual worlds are a simple enhancement of these games, more complex simulations will follow as the technology advances. With this in mind the development of a perfect simulation of our history follows as a natural progression of the development of virtual worlds rather than the initial intent.
As to the creation of a complex virtual simulation of our ancestors, it is nearly impossible to prove we are not living in such a simulation built by a more technologically advanced version of ourselves. Additionally, it is more likely we are inhabiting one of these simulations than experiencing a true Cartesian reality. I have argued it is inevitable that we will eventually develop what Bostrom calls an “ancestor simulation,” – a simulated version of our forbears – and therefore it is nearly impossible to prove we are not currently inhabiting one of these simulations. If we are going to build these simulations, we know it is possible to create a virtual reality which is in itself, real. Not a mere simulacrum, an exact, real copy of the universe and the humans inhabiting it.
In contrast, we can use the same conclusion to argue that it’s impossible to know this reality is a simulation. Descartes himself said he “clearly and distinctly” (Descartes) perceives the world around him, and therefore – ignoring the need for a guarantor – knows what he is experiencing is real.
Nevertheless, even if our reality is not an exact replica of the reality on which it is based I’m not convinced any differences could actually matter. Nor should the idea of inhabiting a simulation cause us to question our knowledge or the nature of our being. This is all we ever will know, and all we ever can know – the distinction between real and simulated is immaterial. Undoubtedly, we will eventually build a simulation that matches the complexity of this reality, and there is no practical way to prove we are not already existing within such a simulation, regardless of any clear and distinct perceptions or other methods of rationalization.
However, there is always the possibility that we will be the first human civilization to build such a simulation, but this is less likely than my initial assertion. Bostrom’s paper goes into the minute detail of why this is so, and I will attempt to summarize. To illustrate, we must first take into account all the possible outcomes for the human race that do not end with self-annihilation, a surprisingly small number, but by looking at human history (nuclear weapons, biological warfare, etc.) it is easy understand how this factors in. As we haven’t yet annihilated ourselves, and are on the veritable cusp of what some philosophers call the post-human state, we are left with two possibilities: we are living inside a simulation developed by an advanced civilization of humans, or we are going to annihilate ourselves before we build our own version of said simulation. The reasoning is thus: the likelihood of self-annihilation vastly outweighs the possibility of the development of a post-human civilization, therefore if we manage to build our own ancestor simulations we are most likely living in a previous iteration of that same simulation. To put it another way, we are following the only successful path to a post-human society because it is the only path available (Bostrom). And yet this is not to say that we are living in some fated, pre-determined world, simply acting out what has already happened, because all human societies will either reach the post-human state, or in the process annihilate themselves. Therefore, it is highly unlikely we will be the first generation of humans to build our own version of the ancestor simulation.
If this is true, if we are only a series of computer programs running in some vast post-human server farm, why should we continue to exist? The human experience is too complicated to be captured in a computer program, no matter how complex and true-to-life. Furthermore, I am the sum of my experiences, not the result of a complicated algorithm. Simulated personalities can’t account for the variation among individuals, a simulation couldn’t generate so many completely unique variants.
Though in a traditional sense this is true, there is no way we could simulate a human being with the technology of today. One must take into account that the technology of the future will be advanced beyond our current comprehension; a post-human world will be as different to us as our world would be to pre-historic man. Specifically, we will not be building programs to simulate a human consciousness, we will build what now, in it’s infancy, is called a virtual neural network (or something using similar principles.) A virtual neural network is essentially the wired equivalent of a human brain; an exact neuron to neuron copy of the structures and systems of the mind. There is no need to program such a machine, it could be born, live, and die without any outside intervention. To explain, these neural networks operate and learn using the same principals as human biology, they are not programmed, instead learning in the same way a person learns – by experiencing the world around them.
I almost hesitate referring to such a device as a machine, as it acts on the same principals as the biological brain. To demonstrate, our brains experience the world through our senses, not through direct contact with the world around us. We use our senses to interact with the world and our minds interprets these interactions. Therefore, there is no reason why we couldn’t exactly simulate a conscious system with technology.
In conclusion, Descartes was able to move beyond his conclusion that he was living in a dream, but I have argued against this. I have argued that it is more likely we are living in a simulated reality than a Cartesian reality, and Descartes’ expedition into the basis of knowledge went a few steps too far. I have demonstrated that humanity is likely to develop our own simulated realities, and that we will use these virtual worlds to create reflections of ourselves and eventually simulations of our ancestors. This is not only possible, it is probable. Furthermore, due to this probability I have concluded it is more likely we are currently inhabiting an older iteration of one of these ancestor simulations than a Cartesian reality. There is no need for our reality to be real in a Cartesian sense, only that we are able to clearly and distinctly perceive it. And yet, even when the vain or overly complex nature of my argument is called into question the fact remains; humans are currently in the process of developing simulated realities at this time, and through an understanding of the rapid evolution of technological complexity these objections become specious. In summary, Descartes needed God, I take more comfort in humanity, but regardless we can all take comfort and enjoy another day in all the wonderful complexity of our shared reality.
Bostrom, Nick. “Are You Living in a Simulation?” Philosophical Quarterly 53.211 (2001): 243–255. Print. Link: https://www.fat.bme.hu/student/pub/Programozas3/SimulationArgument.pdf
Descartes, Rene. “First Meditation: What Can Be Called into Doubt.” Descartes: Meditations of First Philosophy. Ed. John Cottingham. Cambridge University Press, 1996. 12–15. Print.