Solving the Riddle of AI
Computers can be taught to play chess, drive cars, recognize faces, understand speech, and talk to us. How did we endow computers with this level of artificial intelligence? And how much further can we go in getting them to reason like human beings, think for themselves, and demonstrate common sense?
Solving the Riddle of AI
Researchers turn to three main sources for clues about how to build intelligent machines: philosophy and logic (thinking, abstractly, about thinking); life (studying smart animals, including us); and engineering (examining intelligent behavior in the machines we build).
All three approaches have led to advances, and all three are still actively pursued.
The artificial intelligence problem is taken to be that of making a machine behave in ways that would be called intelligent if a human were so behaving.
The roots of MIT’s CSAIL go back to the AI lab founded in 1959 by John McCarthy and Marvin Minsky.View Artifact Detail
Creating the Tools for AI
The quest for Artificial Intelligence, begun in computer science’s infancy, stimulated new software languages, algorithms, and computer hardware.
The programming language LISP (“list processing”), invented in 1958 by John McCarthy at MIT, became the “lingua franca” for AI. Many AI algorithms became a standard part of the computer scientist’s toolbox.
LISP, the preferred language for AI, ran slowly on expensive conventional computers. This specialized LISP computer, both faster and cheaper, was based on the CADR machine designed at MIT by Richard Greenblatt and Thomas Knight.View Artifact Detail
Symbolics emerged from MIT’s AI Lab and in 1981 began making LISP machines for AI research and commercial applications. The 3600 series also supported high-resolution color graphics and had sophisticated movie animation software.View Artifact Detail
Japan encouraged research in artificial intelligence based on the Prolog programming language. This personal computer for AI education came with voice recognition software and a touch pad.View Artifact Detail
Who Studied AI?
Universities led early AI research: Carnegie Mellon, MIT, and Stanford, later joined by the University of Edinburgh in Scotland. Much early funding came from governments, particularly the military.
Starting in the 1980s, the Japanese government invested heavily in a largely unsuccessful “Fifth Generation Computer” project focused on AI and parallel computing.