According to this article, artificial intelligence (AI) can be defined as, “a sub-field of computer science” with the goal of developing computers that can learn and behave intelligently. That description leaves a very wide range of what constitutes an artificial intelligence, and as the article states, many people in the artificial intelligence community fall into one of three camps.
From the same article: “For some, the goal is to build systems that think exactly the same way that people do. Others just want to get the job done and don’t care if the computation has anything to do with human thought. And some are in-between, using human reasoning as a model that can inform and inspire but not as the final target for imitation.”
The first camp is referred to as the “strong AI” camp, the second being “weak AI”, and the third is somewhere in between the two. One important thing about the third, in between camp is that they seek to use the best parts of human intelligence and learning as a starting point for developing their systems. Arguably, these systems are the most powerful and advanced of the three camps, since theoretically they have the potential to surpass human intelligence.
Each of the three disciplines of artificial intelligence is remarkably similar to human intelligence since we use our own learning schemes and logic as a baseline and starting point for developing them. However, computers have the capacity to be much better at learning than humans. As pointed out by this article, human brains are incredibly complex and are designed to do much more than just learn. In addition, humans are inherently imperfect, whereas computers can repeat the same action or process thousands of times perfectly each time. Thus, artificial intelligence is similar to human intelligence in that we can apply the definition of intelligence to it, but it is not limited by that definition and can surpass us.
I would say that existing artificial intelligences such as AlphaGo, Deep Blue, and Watson are proof that artificial intelligence is viable and can be achieved. Watson, for example, uses the vast amounts of information available to it on the internet to analyze patterns and find evidence in order to support a claim or idea. It uses this evidence to establish a certain level of confidence in its answer. This is a basic example of Watson can do, but it is a proof of concept that we can develop systems that can, at the very least, imitate intelligent reasoning. We are knocking at the door of creating a system capable of real intelligence.
So is the Turing Test a valid measure of intelligence for AI? Well according to the article here, passing the Turing Test does not always indicate that the AI is truly intelligent. As the article goes on to state, some systems can just imitate intelligence. The Chinese Room provides a good example of this. If we use the total internet as the books in the room, it isn’t unreasonable to think that a system like Watson could search for the correct response in that context and output it based on what it found. I believe that in order to determine if an AI is actually intelligent, we will need to watch its inner workings and not just the outputs that the Turing Test observes.
Can a computing system be considered a mind? Maybe. We just don’t know. In reality, the study and development of artificial intelligence is still in its infancy. We don’t know where it will lead and what can come of it. Humans by our understanding are more than just biological computers in some senses, but if you believe in evolution then we are really little more than just a system designed to survive long enough to reproduce. We have developed abstract ideas as a result of emotion and thought such as ethics, morality, and intuition that guide us to make sometimes irrational decisions. We are in a sense, just biological computers with a specific task, but to ourselves and each other humans are much more than that. However, the implication that a computing system could be considered a mind means the ethics we apply to each other are the same we need to apply to apply to those systems. We have laws about the way we treat animals, even the ones that exist almost exclusively as a source of food. There would need to be some serious debate on the ethics behind creating these minds, what they were created for, and what the circumstances would justify “killing” or destroying that system.
ethics101withkylewordpresscom 