Artificial Intelligence and Consciousness

The problem of consciousness is the main issue in the philosophy and science of mind. How do brain states, neural activity generate consciousness? Attached to this problem, like a rider on a legislative bill, rest the question of artificial intelligence, which is essentially concerned with computational intelligence.

Although John Searle seemed to have put the issue of artificial intelligence (AI) to rest years ago with his brilliant “Chinese Room” example, there are still those who believe that machines will leave us behind mentally, perhaps supplanting us altogether?

Given all the research on programs that can adapt and learn and can be said to have some type of intelligence, we need to examine our definitions to avoid a continual debate.

Artificial is usually defined as “made or contrived by art, or by human skill or labor,” or “made in imitation of something natural; simulated.”  While these definitions seem almost the same, they imply subtle differences. The first definition is value neutral. It simply states that a thing is made, rather than occurring in nature. The second definition includes words that are not value neutral: “imitation” and “simulated.”  An imitation is usually not as good as the original, but if it turns out to be better, than the original is suddenly seen as flawed. 

Intelligence is an even more loaded word. “The ability to learn or understand from experience; ability to acquire and retain knowledge; mental ability. The ability to respond successfully to a new situation, use of the faculty of reason in solving problems, directing conduct effectively. (Psychology) Measured success in using these abilities to perform certain tasks.”

Intelligence also has the military meaning of gathering and processing information.  

At this point things get ambiguous. Computers are very good at gathering and processing information, and to some degree they can accomplish some of the other parts of the definition. They can learn and retain, and they can respond to new situations, and they can use these to perform certain tasks. However issues of mental ability and the faculty of reason are the sticking points. Can machines do these things or will they ever be able to do them?

Mental abilities and reason are usually considered traits associated with an intelligence that is self directed. The intelligent agent has needs, desires, curiosity, goals, interests, and reasons to seek information. This information, once assimilated and applied to these needs, goals, etc, become knowledge. They become known and can be used to synthesize, analyze, and evaluate. The results are applied, not because the program has an application, but because the intelligent agent has a purpose and an interest in this application.  Something known has qualitative value in addition to quantitative value.