What is Artificial Intelligence?

Alison Adam analyses a type of computer system that comes under the heading of ‘artificial intelligence’ or AI. That is the type of knowing which goes on in the ‘thinking machines’ such as the computer systems, simulations and robots. AI’s difference to other types of computer based systems is that AI "claims to model aspects of human intelligence" (Adam, 1998, p. 2). AI refers to a class of computer system designed to model some aspect of human intelligence, whether it be:

  • learning (machine learning);
  • moving around and interacting in the world (robotics and vision);
  • reasoning towards a solution to a problem (search strategies);
  • using natural language;
  • modelling intelligence according to neural models (neural networks or connectionism) or;
  • having expert knowledge of some subject (expert or knowledge-based systems). (Adam, 1998, p. 1).

It is this last definition of AI that interests Alison Adam. She says:

The idea behind an expert system is that it should be based on the knowledge of one or more experts in a given domain, such as law, engineering, science, medicine, etc. An expert system should be able to dispense expert advice at or near expert level. This means that it should give accurate judgements and it should also give its advice at a similar speed to that of an expert. (Adam, 1998, p. 41)

In her criticism of AI, Adam makes use of two AI systems. The first is the Cyc project — an example of expert system — to demonstrate how it is that feminist epistemology can be used to question this expert system. The Cyc project has been under development since 1984 by Lenat with the aim of building "a vast knowledge base spanning most of human consensus knowledge" (Adam, 1995, p. 360). The second is the Soar system — what Adam calls the brainchild of Allen Newell, John Laird and Paul Rosenbloom — which is a problem solving system that solves problems through "formal descriptions of the whole state of the problem" (Adam, 1998, p. 91). Soar has two main objectives, the first is on search methods for problem solving and the second is its use of rules for representing knowledge (Adam, 1998, p. 92). The emphasis in Soar is not on modeling intelligence through knowledge but rather through the use of architecture as the key element in replicating knowledge. It includes amongst its vast applications that of learning, language processing, and various kinds of tactical air simulation (Adam, 1998, p. 91).

 

~ Intro ~ Alison Adam ~ Artificial Intelligence ~ Adam's views ~ Feminist Epistemology ~ Expert Systems ~

~ Strengths ~ Weaknesses ~ Traditional Epistemology ~ Works Cited ~

 

 

 

 


Valentina Mello Ferreira Pinto
Communication Studies/Humanities Double Major

Communication Studies Program, Social Science Division
York University, Toronto, Ontario, Canada, M3J 1P3