It is a phrase that inspires awe of the accomplishments of the human race. It is a phrase that inspires fear for the mankind. Artificial intelligence has the potential for great, wondrous things, as well as those that are horrid and terrible. Writers have often played on this fear of other forms of intelligence – this fear of the unknown – and nothing can be worse than being outsmarted and surpassed by one’s own creation. Artificial intelligence, or AI, has the potential to be extremely beneficial to humanity, but there are costs and problems that could arise as a result of it.
In truth, AI does not necessarily refer to laser wielding robots hell bent on human eradication. “Artificial” is defined as “made by humans; produced rather than natural” and “intelligence” as “the capacity to acquire and apply knowledge” .1 AI itself is referred to as “…the scientific understanding of the mechanisms underlying thought and intelligent behavior and their embodiment in machines.”2 It is the science and engineering of making machines that are equally or even more capable of achieving goals in the world as are humans. There are many different branches and types of AI, each with its own applications and approaches to problems.
Logical AI: all facts about the situation and the goals are represented in mathematical logical language and the program decides by understanding that certain actions lead to certain results and eventually the goal Search: the program examines a large number of possible actions, understands how to do this more efficiently Pattern Recognition: program makes observations, then compares to patterns Representation: facts about the world are represented in mathematical logic Inference: from some facts, others can be understood Learning from Experience: (based on connectionism and neural nets) programs learn what their “facts and behaviors their can formalisms represent” Planning: given general facts about the world (facts about results of given actions), the situation, and the goal – then form sequence of actions to achieve goal Epistemology: study of the kinds of knowledge that are required to solve problems Ontology: study of exiting objects and their properties Heuristics: way of trying to infer something from within a program (ex: how far from goal) Genetic Programming: technique for solving programs by randomly combining two programs and selecting fittest in each “generation”
Artificial intelligence basically works through one of two methods:4 5
Symbolism or the GOFAI (Good Old Fashioned A I) approach attempts to represent human knowledge and rules about the world as facts and then inputs them into the computer as “symbols”. The symbols can then be manipulated and new ones created. The computer can examine numerous possible manipulations and outcomes. Unfortunately, computer systems such as this lack basic common sense that humans take for granted. The amount of data that...