A Detailed Look at AI and other questions presented by the Matrix
Agent Smith pulls a silver metal chair around and sits down in front of Morpheus. "I'd like to share a revelation that I've had, during my time here. It came to me when I tried to classify your species. I realized that you're not actually mammals," he says slowly and meticulously enunciated. "Every mammal on this planet instinctively develops a natural equilibrium with the surrounding environment, but you humans do not. You move to an area, and you multiply, and multiply, until every natural resource is consumed. The only way you can survive is to spread to another area. There is another organism on this planet that follows the same pattern. A virus." The menacing man in black pauses for a moment, letting the statement sink in. "Human beings are a disease, a cancer of this planet, you are a plague, and we are the cure." He says "we" referring to the race of artificially intelligent entities that now populate the planet we once called home.
the Matrix is a movie about a bleak future, where humans have been forced to live deep within the earth's crust in order to survive the onslaught from the very intelligence they created. It is easily one of the best-realized and flawlessly created futuristic movies since Blade Runner debuted in 1982. However, many, after viewing the Matrix, wonder about the feasibility of many of the things within the Matrix universe; the way in which the "hard-lines" work, how the Agents are integrated into the Matrix's code, and the very AI of the machines in general. How valid is what the writer/directors, the Wachowski brothers, have envisioned?
In this viewer's opinion, extremely. When I first saw the Matrix, I was captivated by it as an action film. I won't go into why, but every time I watched it I slowly realized the genius of the film; religious 1 and literary 2 references, things which seemed so well researched, combined with visual effects which completely suspend the belief that is not reality; which, ironically, is one of the themes of the movie. The purpose of this paper is to explore the feasibility of this movie.
To start off with, let's look at the issue of the hard-lines. Hard-lines were phone wiring which allowed the freedom fighters (i.e., Morpheus, Trinity, et al) to escape out of the world of the Matrix and back into the real world. However, those phone lines were merely part of a computer simulation, therefore, there was nothing physical about them, right? To understand why those lines were accessible, we must consider the Agents themselves. The Agents, Morpheus explains, are "sentient programs. They can move in and out of any software still hardwired to their system." Basically, any human still connected to the Matrix physically could be taken over by an Agent. However, the Agents are not part of the Matrix itself. They are like Java applets, they exist independent of the web browser. Therefore, there must be a location every couple blocks in a city where those Agents can be imported into the Matrix. Basically, the freedom fighters can hack those hard-lines and use the Agent's own entrances to get into and out of the Matrix. So although hard lines are really only present in the software, they are coded to be in/out data paths.
After Neo's initial meeting with Morpheus, he takes a red pill which will "disrupt your input/output carrier signal" so they can locate him. Is it possible to give somebody a pill in the Matrix and use that to track their location in the power plant? Absolutely. Remember, the Matrix takes things that occur inside of it and send that information to the brain via electrical impulses. If the drug he takes has the effect of telling the Matrix to send electrical impulses that disrupt the mechanics of his input-output system, there will most likely be an error message that appears on the machine's monitoring systems reporting this malfunction. If the freedom fighters have hacked that system, they can find which human that is and change his vitals status to 'deceased,' effecting the system to have his body flushed so they can get his body via the sewers. This is realistic, as we know Morpheus's crew is extremely good at hacking.
Looking at the bigger picture, we ask about the possibility of interfacing a human with a complex computer program. According to Eliezer S. Yudkowsky 3, it's probable with the invention of nanotechnology. Nanotechnology is the creation of robots on the atomic level. Robots that are composed atom-by-atom to be as small as a cluster of molecules. These robots could manipulate material at the atomic level themselves. Therefore, a single nanotechnology robot (or nanobot) could build a second nanobot, and those two would make four, and the number would grow exponentially. Once enough are created, these nanobots could reconfigure biological material on the molecular level as well. What do nanobots have to do with hooking our brains to computers? Everything. These nanobots could reconfigure our brains so that the neurons have inputs and outputs, where the data could be carried to and from the Matrix. According to Yudkowsky, "You could even abandon bodies entirely and sit around in a virtual reality environment, chatting with your friends, reading the library of Congress, or eating three thousand tons of potato chips without exploding." Does this sound a bit like the Matrix?
But we wouldn't be asking the previous three questions if creating Artificial Intelligence weren't possible. This topic provides us with probably a greater dilemma than that of human cloning, because it is impossible to control something that is more intelligent than yourself. But we're looking at feasibility, not ethics. To understand if AI is possible, you have to look at the human brain. It's a complex network of billions of neurons, with trillions of inter-neuron connections. According to most estimates, the computing power of the mind is roughly one hundred thousand trillion, or 10^17, operations per second. The software has been created and modified over millions of years by evolution. Computers currently have one-billionth that processing power, and the development of AI software is in its infancy. Moore's Law, which states that processing power will double every eighteen months, has held true for the past 30 years. Following a doubling of processing power every two years, computers should reach this sometime around 2025, sometime sooner if you include multiprocessing. As Morpheus said, "Some time at the end of the 21st century, man was united. We marveled at our own magnificence as we gave birth to AI." That would give all the worlds best programmers, collectively working on writing AI, close to a century to complete the project4. That timeframe for that many people is definitely feasible. The Singularity Institute hopes to have finished work on an initial Seed AI by 2008. Morpheus says the AI was "a singular consciousness that spawned an entire race of machines." Again, quoting Yudkowsky:
For the Minds (AI systems), one year of objective time equals two years of subjective time. And since these Minds are human-equivalent, they will be capable of doing the technological research, figuring out how to speed up computing power. One year later, three years total, the Minds' power doubles again - now the Minds are operating at four times human speed. Six months later... three months later...
Basically, these systems would become smarter at an exponential rate. Using a few data points, we can create a parametric equation that will show how much smarter than humans they will be after a certain number of years.
X = 4 - 4(N-2); Y = 2N (Where X is the number of years, Y is how much smarter than us they are)
Taking these two parametric equations and combining 5 them to eliminate N eventually yields this equation:
Y = 4/(4-X)
This equation shows that the machines could begin to possess the intelligence to completely understand things we as humans have yet to comprehend, such as how the brain functions, the exact process of human growth during gestation, new form of power generation, etc., after a mere four years. Essential, they would be intelligent enough to clone us, insert I/O systems into us as we're developing, and hook us up into a highly complex powerplant to fuel themselves. Given, this assumes that after transistor-based computing reaches the point of dimishing returns, other computing paradigms exist to continue the exponential growth. Raymond Kurzweil is fond of pointing out that Moore's Law is the 5th iteration of what he calls the "exponential growth of computing," and that each time the prior 4 technologies have reached the upper limits of their speed, the next paradigm was there in time to allow the growth to continue unfettered. That has been the trend since the first computer was constructed in 1890, so it's safe to say this will continue.
Looking at the sheer possibility of each of these things, one may begin to wonder, "Is the Matrix going to be what happens to the human race?" Unfortunately, we don't know. Hate, envy, jealousy, etc., are all human developed emotions that evolved over hundreds of thousands of years. Assuming an artificial intelligence will possess flaws such as The Devil's contract problem is a broad assertion. There is no guarantee that a proper AI would possess any anthropomorphic characteristics. However, the Matrix does appear to be a possible outcome of a Failure of Friendliness scenario caused by faulty Friendly AI. There is no denying that nanotechnology, AI, and cloning will be realized much sooner than the end of the 21st century, but as Neo says to the AI at the end of the Matrix, "I don't know the future. I didn't come here to tell you how this is going to end. I came here to tell how it's going to begin."
- Many have described the Matrix as a "Cyberpunk Parable," one which sets Neo as a Christ figure with other characters representing other biblical figures. For a more in-depth analysis of the religious symbolism, go to:
- Literary references are detailed thoroughly at a page which no longer seems to exist. You can see google's cache of said page here:
- For more golden nuggets of insight by Mr. Yudkowsky, visit the Low Beyond at:
- Without a marketing department it's amazing what kind of work can be accomplished.