Artificial Intelligence, also known as A.I., is a breakthrough concept that’s already affecting the world around us today. Many of you have seen this in movies, such as Star Wars (the droids) or A.I. (all about Artificial Intelligence), and might be wondering just when we’re going to reach that stage of robotic development. Well, here’s some info you might like. One note - there will be no schematics or detailed information regarding certain aspects as of yet due to the secrecy involved in research.

The University of Alberta is one of the many universities and research organizations currently investing its time and resources into the advancement of Artificial Intelligence. They have broken down the development process into 7 basic areas; Vision, Learning, Cognitive Agents, Natural Language, Games, Constraints, and Representation and Reasoning.

Vision is vital to every single aspect of normal human life. We are visual people and have thusly designed the world around us in that manner (why you’ll see a woman who is…well endowed on the cover of a magazine, not an extremely stimulating cover gloss) so the university has deduced that these robots with artificial intelligence must be able to recognize the world around them visually and react to it thusly. But not only are they sticking on advanced mechanized eyeballs; they are also trying to add reasoning to the sight process. They have a research field dedicated to the prediction of events. One might associate this with instincts, i.e. if a man is running after you get out of the way.

Learning is by far the most important aspect of research be doing with Artificial Intelligence. Most people can’t even teach their dog to roll over, and yet if Artificial Intelligence is going to work we need to be able to program a computer to learn. There have been some advances in this field of programmed learning, but it is very limited compared to the scale required for a working self-sufficient intelligent robot. As is they are trying to devise ways to make the intelligent robots learn by example and repetition. Not only repetition is being used as a means though. Also being explored is the idea of similarities. We humans were taught on one computer. But we switch makes and models almost on a daily basis without the need to relearn everything regarding the concept of a computer. They are also trying to devise similar learning methods with these intelligent robots. The hardest aspect will be programming in curiosity, which is something that comes naturally to us. Getting a computer to learn something isn’t exceedingly difficult for we can already teach them some limited things and with more effort could literally teach them everything. But having a computer completely self-sufficient would mean that it would have to create it’s own theories and test them. And, as with everything, it’s not that easy (and I use the term easy loosely). Some theories and curiosities could literally take a computer till the end of the universe to figure out. So there must also be some sort of variable time limit established with theories.

Unfortunately, we will not be able to order around a self-sufficient, intelligent, and sentient robot in the same way that we now order around mindless, helpless robots.

The area of Cognitive Agents is quite broad, but ultimately it deals with reasoning. The ability to reason is an attribute that only humans posses."

No other creature has it. Reasoning perhaps can best be described as the ability to overcome one’s instincts, or agree with them. Animals rely purely on instincts. If you run after a rabbit the rabbit will hop away. Whereas with humans just look at football. Even though your instincts tell you to get out of the way of the 300lb organic freight-train about to run you over, we posses the ability to stand there and ‘hold our ground’. The problem involved here is that it is easy to program computer instincts. It is also easy to override them. The question is how do we allow them to override their own instincts and react based upon their given situation. This is the area in which intelligent robots could far surpass us. With the ability to calculate the force an object has coming at it, a robot can accurately deduce whether it can withstand the assault with ample time for the assault occurs. Humans can’t and that is why sometimes we make really dumb decisions.

The Games aspect that the university is researching is actually a very old area of computer development. Anytime you play the ‘computer’ in an electronic game you’re actually playing artificial intelligence. What the computer does is it runs billions of calculations ultimately determining the best outcome for itself. Such as when Deep Blue beat Kasporav, the Russian Chess Champion, it had to perform calculations with moves, then calculations upon future possible moves with each move, then make calculations about future possible moves from those possible moves, and so on until it has satisfied a victory only situation for itself. This process requires massive amounts of processing capability. So the only aspect that is really being researched under this is the advancement of processors.

Natural language is pretty self-explanatory. Are you familiar with voice recognition? Or perhaps writing recognition in PDA’s? These are both forms of natural human language. Right now, they are trying to make computers not only associate these squiggles and sounds with meaning, but also to allow the computer to understand these squiggles and sounds.

That about sums up all the efforts being done not only by the University of Alberta, but also every other research group in the field of AI. Now all we have to do is sit back and wait for servant robots to takeover all aspects of work in our day to day lives…or not.

Article written by Zachary Skelly.