GOFAI

GOFAI (Good Old Fashion Artificial Intelligence) is narrowly defined as those AI theories that adhere to the tenets of Hobbes' "rationalization is computation". This means that any theory of AI that satisfies GOFAI must contain some component that handles the individual manipulations that compromise intelligent action. GOFAI is totally invested in knowing that as opposed to knowing how.

GOFAI (Good Old Fashioned Artificial Intelligence) is a term coined by John Haugeland in Artificial Intelligence: The Very Idea (1985). In simple terms, this refers to an artificial intelligence that works by taking in a rule set and outputting working solutions.

This includes programs that try to brute-force calculate the optimal moves in games of checkers and chess and the expert systems of the 1980s. These systems were very 'brittle', meaning that even a small error in input could derail them, and they learned by being given new rules or facts, rather than deducing new rules or facts. This sort of AI is fairly weak by today's standards, but it is important to note that GOFAI is what most people think of when they think of an AI, e.g. Deep Blue and ELIZA.

Modern AI systems are enhanced by artificial neural networks, evolutionary computation, and deep learning, among other techniques.