Philosophical Cud: The Simulation Hypothesis

Hello there. Before we get down to business, I’d like to take a moment to explain who I am and what this column wants to do.

I’m a senior here at Luther, and spend most of my time in the philosophy and computer science departments. Philosophy is where it’s at for me, but on the chance that my academic pursuits don’t live to fruition, computer science will – hopefully – ensure that I remain employable. That, and displaying my most recent software projects to the parents is usually a more effective means of showing what I’ve been doing with their money than explaining what Nietzsche or Foucault thinks about the human experience. With that said, I’d like to use this page to give all of you folks something to think about while you’re taking some downtime with Chips. My goal here is to take philosophical theories as well as less formal ideas and make them more palatable for leisure reading. I’m doing half of the digestion myself, and returning something thought provoking that can be taken in on a coffee break – thus the title, Philosophical Cud.

Sooner or later, anyone who cares to ponder how material bodies interact with self-conscious minds will realize that there is no good way to definitively know anything about what really goes on outside of our heads. As it turns out, it’s impossible for one to know that she isn’t a brain floating in a vat, being fed electrical signals which emulate the sensory input we all experience every moment. Many of us felt the anxiety that is usually associated with this dilemma after seeing The Matrix for the first time. Nobody wants to be milked by machines, but what if it gets worse? What if our brains weren’t merely hooked up to a device that simulated reality, but instead our consciousnesses were themselves artificially intelligent simulations? Dr. Nick Bostrom of Oxford theorizes that this scenario isn’t only possible, but may in fact be probable. Enter the Simulation argument.

At the heart of the Simulation argument lies the claim that one of the following three propositions must be true.

The first proposition asserts that almost every species becomes extinct before reaching a level of technological advancement that allows for the creation of artificial intelligence.

The second proposition is that nearly every species that does attain this level of technological prowess chooses not to create artificial intelligence.

Finally, we ourselves are nearly certainly artificially intelligent simulations. Let me explain what these statements mean, and why there might just be reason to take them seriously.

First, consider a future in which the human race had advanced to a technological state which allowed for the development of artificial intelligences.

Consider further that at the point humans became capable of rendering consciousness, they would also have access to much larger amounts of energy than is currently producible. (For more on the relationship between technological progress and energy consumption, look up the Kardashev scale.)

If future humans decided to create artificial minds, they would have the computing resources to to simulate more consciousnesses than have ever naturally existed. As time went on and simulations continued to be run, the disparity between the number of natural consciousnesses that had existed through history and the number of human-designed ones would broaden until the majority of consciousnesses that had ever existed were those which had been implemented technologically.

This projection illustrates why one of the three initial propositions must be true. Either a civilization dies off before it can simulate intelligence, chooses not to, or it becomes a matter of time before there have been more ‘fake’ minds than natural ones. The Simulation argument makes no bets as to which of these three propositions is in fact true, but I find the case for the third option to be rather plausible – assuming that some type of artificial intelligence is in fact engineerable. If we can manage to both keep from blowing ourselves up and to discover artificial intelligence, I’m sure it’s the case that intellectual curiosity will make research into mass simulation happen.

So, then, it’s at least theoretically plausible that if a species manages to develop to a point where it can generate cognizance, and generates it long enough, it’s probable that we, being conscious beings, are in fact merely artificial intelligences experiencing a simulated world. There is, of course, no evidence for this whatsoever, but the implications of such a reality may be worth thinking about. If it turns out that you are indeed a man-made consciousness, it means that you weren’t created in God’s image, you were developed in the image of a less than divine engineer. As a manufactured, disembodied mind, do you have a soul? If the simulation is maintained, it’s possible that you may live for thousands of years, if not forever. Perhaps you’ve only been alive for a moment. Do rational productions have moral responsibilities? While you won’t ever know if you’re a simulated intellect, the implications of existence in such a state are staggering.

Does the simulation argument provide a valid philosophical picture of the universe, or is it merely science fiction? Before making up your mind, you may want to read Bostrom’s original article, which can be found at www.simulation-argument.com. I’ll let you chew it over. Until next week, stay out of trouble.