In the Mind of the Machine

by Ben

April 29, 2005

Today's modern computers are one dimensional. They serve the purpose of running the programs we want to run, and that's about it. They have no intelligence, no opinions and no common sense. It's no surprise then when programs crash, because the machine doesn't know any better than to realize its riding on a bike with a flat tire. In the future, this may no longer be the case. Putting aside the idea of complex Artificial Intelligence, computers will have to be more capable of logic, not just binary logic. A yes or no answer may not always provide the desired results.

In mecha anime, we've seen many ways that machines have been given life, or at the very least, a sort of automatic pilot, that allows the machine to perform without a pilot, but what actually happens? In some futures, the development of artificial intelligence has come far enough that the machine does think, but without emotion. Because of that, the machine might make cold or irrational decisions in our mind, that might be completely logical to it. If your car was barreling out of control, and your only options were to hit a pedestrian or hit another car, most people would opt for the car, since it would most likely absorb a lot more of the impact, since the pedestrian was more likely to be killed on impact. In a more focused logic, not based on that emotion based decision, your car might opt to hit the pedestrian, since it would more than likely protect the occupants of the vehicle more. Another alternative means of giving intelligence to a machine or a construct is the implantation of a soul or essence of some person. Since in reality, we barely understand the realities of how our brain and body work, its very unlikely that the essence of people could be transported into the memory or logic of a machine.

Yet another alternative choice is a combination of pilot and machine. In Gundam Wing, we see the Zero system, a complicated artificial system that alone is incapable of action, but by interfacing with a pilot can give more or less suggestions on combat to the pilot based on facts, and maybe a little bit of premonition. This type of system can overpower the normally logical decisions of the pilot by implanting confusing information, and could be more dangerous than good.

Now all of this isn't to say machines in the future won't malfunction, no matter how complex or how many safeguards are in place. We ourselves are equipped with very complex computers (our brains) and we still malfunction or come out a lemon from time to time, but the fact is that machines are reaching a plateau in modern development. There is only so much more you can do with an interface that doesn't interact. We can make mock interaction already such as with Chat Bots, but the response is extremely limited. So what is being done to move beyond this barrier of the unthinking computer? Current development is slow at best. It's unlikely we will see a HAL 9000 in our lifetime, and even more unlikely to be walking talking, thinking giant robots. There will be no Autobots for us to hitch a ride from in the near future, but there is hope.

If you visit the website of the University of Florida you can read an article from last October about the development of literally, a brain in a dish. By taking cultured cells from a rat's brain, a UF scientist was able to directly link them to a computer and over time, the brain taught itself how to fly a flight simulator. Not too shabby for a rat brain in a dish. The development potential for this is astonishing. No, this brain isn't going to sign onto aim and start learning English. It's a rat brain for one. But the fact that in a small dish is a learning computer, something that we have not yet been able to even consider in silicon, is incredible. If such a project continues to be productive, the ideas behind brains or *souls* in machines may not be far off. Robocop would be within the realm of possibility. In the modern Transformers mythos, each Transformer is host to a spark of life. Perhaps that spark could be seen as the one part of them that is a living, non mechanical component. Without it, they are no different than construction equipment.

The possible military application for this is also impressive. Instead of equipping cruise missiles with expensive and complicated sensory equipment to detect a target, a living computer could be linked directly to an infrared or UV camera and make adjustments to the location of a target. "Hey! The terrorist is in a jeep heading west at heading 355.4" Suddenly, a cruise missile receives this message and using its navigational systems turns 180 degrees and searches out and finds the target. Living computers are the concept that may lead us into the future. Where programming fails, natural growth may triumph, and maybe a combination of binary logic and empathic logic is exactly what we need for the next computing revolution.

<<back to Mecha Musings