Even though it’s still not (yet) on the list of words with daily use, the concept of Artificial Intelligence (AI) is known by the vast majority of population, and it’s becoming more and more used by the social media and various scientific communities. However, if inquired, most people can only say that this term refers to “something related with computers”.
Evidently, AI is much more than that. It can be seen as a science which as the goal to understand how human knowledge functions and what is intelligence, or it can also be perceived as an engineering, since it works on the construction of gadgets that help and even mimic the exercise of the human mind. To replicate the thought into computational processes, the thoughts are tried to be explained through a set of algorithms. An algorithm must be perceived here as a finite set of instructions, clearly defined and not ambiguous, that can be executed mechanically during a period of time and with an amount of effort that are also limited.
AI is closely related to Cybernetics, they are interdependent, since is Cybernetics that materializes and applies AI theories.
There are two types of Artificial Intelligence, Strong and Weak.
Strong AI is self-conscious, meaning it is aware of itself, and it’s the one who raises more ethical conflicts, since it embodies more imprecise concepts, such as conscience, because through Strong AI it is intended that a machine should be able to think and solve problems, just like a human being, and this could mean that a so called computational system would also feel pain and pleasure, and have its own beliefs and values. The Strong AI thesis, or cognitivism, assumes the difference between hardware and software, and defends that in order to create artificial intelligence, it is the software that must be reproduced. According to this theory, some authors argue that, sooner or later, it will be possible to replicate, neuron by neuron, the human mind on a machine (hardware), which brings up the idea that this way a computer will be able to develop new algorithms by itself, based on rules and metarules, and no longer needing any human to program it.
On the other hand, Weak AI is only a simulation of the human intelligence, the computer would only replicate an intelligent behaviors, but with wouldn’t have conscious. The Weak AI thesis, or connectionism, admits that in order to create AI, it is necessary to reproduce the hardware, even though some elements of the software are still necessary; this theory defends that intelligence can be simulated, but not reproduced. The ultimate goal for the creation of AI would then be machines that show an intelligent behavior, without really possessing it.
It is fair to say that Artificial Intelligence is still on an early stage of its development; a lot of critics and tests have been faced by both forms of AI, and the developments in this science are growing each day.