After taking some time and actually looking into what Cyc and OpenCyc provide, I have a couple quick points to add to my last post. I hate adding content to something already posted (editing for spelling/grammar, etc is one thing, adding or removing content, I try to avoid).

First, Cyc and OpenCyc are both projects of Cycorp. From what I gather from their FAQ's and other information on the project, it is a combination of a reasoning engine, and a large, comprehensive database of terms, and assertions that link the terms together. Now, to come back to what I was mentioning earlier about Artificial vs Synthetic Intelligence, Cyc is a good example, in my opinion, of an usable artificial intelligence. It seems to provide all the needed elements to fake intelligence and allows us to communicate with a machine. What I don't see evidence of, however, is the ability for this intelligence to grown on it's own. (Please correct me, if i am mistaken here).

We will explore a tangent for a moment. Stepping away from the process of creating intelligence, lets examine how our minds work. The human brain (and, as a result, the body as well), is powered by a mix of chemicals and electricity. We have millions, if not billions of neurons and synapses and axons in our heads that provide means of survival. These connections in our head allow us to think, dream, and operate. We still know very little about exactly how the brain works, but that isn't really all that important to develop and S.I., or even an A.I. for that matter.

Beyond the physical make up of the brain; over the years, we have developed an understanding of how we think and how we operate from psychological and philosophical points of view. Our actions are driven by a number of things. First, we have pure instinct. On the most basic and primal level, our instinct is survival. We have drives of survival. This is where the desire for reproduction, and the fight or flight instincts come from. On the most basic level, we want to survive. Next, we can add motivation and drive to the equation. Basically, we all have goals, and want something from life. Before everyone starts poking holes in that statement, I am being very general, and ignoring disorders and the pieces of the puzzle that make all of us unique. Now, back to the blog. There are various things that can and do motivate us. Together, with our primal instincts, we are given direction. But we're still missing something. We need knowledge. What do we need to do in order to accomplish the goals we are motivated towards? How do we get from point A to point B? With the added piece of knowledge, and experience, we have the foundation for intelligence.

Now, it's nearly 3am, and I'm writing this. I know the previous paragraph needs some work, and I do plan on posting something that's laid out more precisely later, but that might be a few weeks away. The structure Andrew and I are using still requires form refining to be usable and accurate to the point that it will work. And then I need to explain it simple terms. But essentially, what we're looking at is 3 points of input: motivation and drive, knowledge and past experience, and something I left out from above, actual thought and imagination. Knowledge and motivation will only go so far. Alone that cannot really accomplish anything. They need a catalyst to give them a direction to go. And that is where imagination, or simply thought comes in. By using these relatively basic concepts, we're able to actually get something accomplished.

Cyc provides the database of knowledge, and in some ways, experience, but in the way that is useful for recreating intelligence. Cyc also provides a way to examine said knowledge and with whatever input you give it to get something useful in return. But again, as I said earlier, this is not really intelligence. It's faked.

However, it is important to note that though Cyc won't create a true intelligent machine, it is an important project and does provide some very useful features and services. And, in fact, the opencyc database, may very well provide a key piece in creating a true intelligence by providing the other pieces with a very comprehensive knowledge base of terms and meanings. As a result, you're A.I./S.I. will essentially be a living dictionary or possibly, even encyclopedia. Together with experience from whatever application is being used for input (and output for that matter), the A.I. may be able to actually spawn conscience and become a real intelligent machine.

---

Man I love this stuff. So thought provoking, and incredibly difficult to develop and maintain. I think it's possible though. And I think we can do it. There a few pieces missing from the puzzle though, but they have to deal with more human concepts that will be easier to explore implementation once the system is in place. Namely, morals, laws, and ethics. How will a machine handle morals? Or respond to legality or his/her or others actions? Or for that matter, handle the area of ethics. My thoughts lean mostly towards a set of rules in your database of knowledge, but I'm not confident that the database will necessarily be flexible enough for such aspects.

--nullpuppy(out)