Sunday, May 4, 2014

AI: Artificial Intelligence

When I was in college, I was learning about artificial intelligence.  But what does "artificial intelligence" refer to?  For some, it means something that is not "alive" being able to learn, while for others, it means replicating human cognitive abilities such as critical thinking and useful data processing.  Intelligence simply refers to the ability to gather and use information.  So given this definition, how do we as humans know that we are natural intelligence (NI), if nobody really understands how we can think?  The idea that intelligence is the effect of chemical signals to me is preposterous because that would mean there is some type of mapping between the neural signal pattern and thought.  For this to be true, there would have to be a built-in design that correlates what is outside of the human body with these mental patterns.  One might argue that the learning component is what creates the neural patterns from the sensory data of the external world, but that still would not explain how the mapping of these patterns represents concepts or thought, especially inner thought or knowledge that had no external data such as for the blind or deaf.  In fact, I would argue that we as physical beings are really artificial intelligent beings because natural intelligence creates information, while artificial intelligence requires information.

When you read about computer scientists trying to create artificial intelligence, they rarely explain how they plan to do this, but there is some complex design behind the technology.  Some work with software, while others work with hardware.  There are many different ways to approach codifying intelligent software, but at the end of the day, these are all designs of humans.  From this point of view, whatever the artificial intelligent thing that is created, is simply the transference of human intelligence into that thing.  Even if someone were to create a system that could learn and "think", it was the creation of a human being.  Even if the created system was sophisticated enough to be able to create another artificial intelligent thing in the future, the original "creator" was a human and it will be limited by the sum total knowledge of the human creators. This is in reference to their behavior, not the data that can be collected. An AI system can only do what it is programmed to do.

So given this notion, what if humans are merely that artificial intelligent "thing" that is able to create other artificial intelligent things in the future?  That would explain why we must learn, rather than demonstrating innate intelligence all the time, because humans use a system of intelligence with learning as the foundation.  Learning through life seems like a natural concept, having just been born into this world.  Because of this, people use reason to justify not having any information at birth, and so we assume we learn over the growth period.  But again, from an artificial intelligence point of view, this is merely a mechanism used to gain intelligence.  There is nothing that says information must be learned.  We have created intelligent systems that already has information built-in and they are called expert systems.  These types of intelligent systems are designed for specific purposes.

From an expert system point of view, you could say animals, including humans, use this mechanism.  They are born with instincts.  These instincts are the built-in intelligent mechanisms that allow them to survive.  Again, from a system design perspective, expert systems are created by intelligent beings that already understand and know the information being stored.  So instincts must also be codified information into an expert system.

In a computer-based AI system, the abilities and information are coded into a program.  For living things, that code is believed to be our genes.  If learning and instincts are built-in intelligent mechanisms, how did these get codified into our genetics?  Where does intelligence come from that allows this type of behavior?  Now the religions all point to a God figure that created these systems.  But the problem with this is that it does not tell the full story about the creation.  This type of story infers that the created things are somehow different than the creator, without even defining who/what the creator is.  I've posted many times my explanation about God and who the creator really is, but from an AI point of view, the creator is the natural intelligence that created the Artificial Intelligence.  The AI is simply a limited set of knowledge and understanding with the ability to learn.  The AI can only be created by a superset of intelligence that creates intelligence.  It is the source of all information and knowledge.

The NI does not need to learn.  It already knows.  It does not need to understand information as it creates information.  It created the concepts.  The only thing the NI does not have is experiencing that knowledge.  It's the equivalent of thinking about something, but never having actualized that experience.  Because of this, the NI created a mechanism to experience the concepts of knowledge as actualized experience.  Just using NI would mean there is no sense of experiencing because it would already understand the nuances.  So in order to experience from a different perspective, it needed to "dumb down" the ability to know.  AI is the mechanism used to experience without knowing, hence the need to learn.  It is coded with simple abilities, but it is not limited to a set of knowledge.  I can learn and grow that knowledge.  Through this mechanism, gaining knowledge is the experience of understanding knowledge.  Just having knowledge is not enough.  The NI wants to experience that knowledge in as many possible ways as available.  Having knowledge and feeling knowledge together is more powerful than either alone.  So the AI that is created was a means to have that experience.

Unlike the AI humans create for whatever purpose that exists, the human AI is a vessel that is designed to carry experience.  Information already exists and is known, which is why we are able to learn and think at all, but the experience of that information is new every time in each body.  The body is the robot that the NI uses to experience, and the AI is designed to gain it.  Without the proxy AI with its own free-will and choices, which creates unpredictability, the NI could not gain the full experience since it would already know all outcomes since it can see all probabilities. But through the AI, the NI can "surprise" itself by observing which probabilities the AI will choose.

Why are we as humans so limited in our understanding, abilities, and capabilities?  Because we are both the NI and the AI, and we designed this body this way in order to have a full experience with the feelings of surprise and the unexpected.  If the human body was immortal and all-knowing, you would be bored after a period of time of the human experience because there would be nothing new.  So you as an formless immortal being with natural intelligence, creates forms to experience through "AI's", or "limited" beings, in every way you can imagine.

No comments: