The mathematician John McCarthy is often credited with coining the term “Artificial Intelligence”. He described the problem of Artificial Intelligence as “that of making a machine behave in ways that would be called intelligent if a human were so behaving.” He used the term in a 1955 research proposal for the Dartmouth Summer Research Conference on Artificial Intelligence which took place the following year. Since the Conference, there have been many comments made by technologists comparing the potential of computers to human abilities. A notable contributor to AI, Herbert Simon, said in 1965, “…machines will be capable, within twenty years, of doing any work a man can do.” Marvin Minsky who was a participant in the original Dartmouth Conference said in 1970, “In from three to eight years we will have a machine with the general intelligence of an average human being.”
The human-comparison—expressed as a research project—comes in two flavors. The first approach has been called the cognitive (also 'psychological') approach to AI. The goal of this approach is to create actual thinking machines. This sentiment was expressed by the philosopher John Haugeland in his 1985 book “Artificial Intelligence: The Very Idea”:
“The fundamental goal of this research is not merely to mimic intelligence or produce some clever fake. Not at all. “AI” wants only the genuine article: machines with minds, in the full and literal sense.”
The second approach has been called the engineering (or technological) approach to AI. This second approach does not claim to literally reproduce human intelligence. Instead it's aim is to use computers to perform tasks with more practical application. Elaine Rich and Kevin Knight wrote in their 1991 book “Artificial Intelligence”, “AI is the study of how to make computers do things at which, at the moment, people are better.” Ray Kurzweil has described AI as “The art of creating machines that perform functions that require intelligence when performed by people.”
According to the philosopher Luciano Floridi, these approaches represent the “two souls of AI” that both claim a single common heritage in the Dartmouth Summer Research Conference on Artificial Intelligence. The cognitive approach to AI attempts to emulate human consciousness. While the engineering approach attempts to only simulate some aspect of intelligent human behavior at best. Strong AI is the hypothesis that the cognitive approach will be successful using computers alone. Weak AI claims that the engineering approach to computer AI is sufficient to simulate intelligent human behavior.
Advances in AI using the engineering approach have been a stunning success. It has given us remarkable technology that makes our lives more convenient such as predicitive typing on our mobile devices. On the other hand, the cognitive approach to AI has been a dismal failure. According to Jerry Kaplan, an AI expert at Stanford University, "Little more than speculation and wishful thinking ties the actual work in AI to the mysterious workings of the human mind.”
The distinction between the engineering and cognitive approaches to AI is not explicit and many suppose that these different research projects are the same. Because of this conflation, it is naturally assumed that progress in one approach means progress in the other, but this is not the case. It is a mistake to assume that there is a natural progression from the current work on computer technology to creating conscious machines. If it is possible to create an artificial conscious machine, a completely different technology than computers will be needed. Trying to use computer-technology to create consciousness will be as fruitful as trying to make a chair out of helium gas. It is simply the wrong approach.
Marvin Minsky quoted in LIFE magazine 1970
John Haugeland, Artificial Intelligence: The Very Idea (MIT Press 1985), 2.
Stuart Russell and Peter Norvig , Artificial Intelligence: A Modern Approach (PearsonEducation, 2003), 2.
Luciano Floridi, The Fourth Revolution: How the Infosphere is Reshaping Human Reality (Oxford University Press, 2014), 140.