The Dartmouth Workshop--as planned and as it happened
Next: Bibliography<br>Up: AI PAST AND FUTURE<br>Previous: AI PAST AND FUTURE
The Dartmouth Workshop--as planned and as it happened
The four organizers of the 1956 Dartmouth Workshop on artificial<br>intelligence were John McCarthy, Marvin Minsky, Nathaniel Rochester,<br>and Claude Shannon. The workshop has some prehistory.
My own interest in AI was triggered by attending the September 1948<br>Hixon Symposium on Cerebral Mechanisms in Behavior held at Caltech<br>where I was starting graduate work in mathematics. At this symposium,<br>the computer and the brain were compared, the comparison being rather<br>theoretical, since the first stored programmed computers were<br>completed only in 1949. The idea of intelligent computer programs<br>isn't in the proceedings of the symposium, although it might have been<br>discussed. I developed some ideas about intelligent finite automata<br>but found them unsatisfactory and didn't publish. Consequently, I<br>wrote my Princeton PhD thesis in differential equations in 1951.
Marvin Minsky was independently interested in what became AI and in<br>1950, while a senior at Harvard built, along with Dean Edmunds, built<br>a simple neural net learning machine. At Princeton, Minsky pursued<br>his interest in AI, and his 1954 PhD thesis established the criterion<br>for a neuron to be a universal computing element.
Claude Shannon proposed a chess program in 1950, built many small relay<br>machines exhibiting some features of intelligence, one being a mouse<br>searching a maze to find a goal.
In the summer of 1952 Shannon supported Minsky and me at Bell<br>Telephone Laboratories. The result of my efforts was a paper on the<br>inversion of functions defined by Turing machines. I was unsatisfied<br>with this approach to AI also.
Also in 1952 Shannon and I invited a number of researchers to<br>contribute to a volume entitled Automata Studies that finally came out<br>in 1956.
I came to Dartmouth College in 1954 and was invited by Nathaniel<br>Rochester of IBM to spend the summer of 1955 in his Information<br>Research Department in Poughkeepsie, NY. Rochester had been the<br>designer of the IBM 701 computer.
While at IBM, Rochester and I got Minsky and Shannon to join us in<br>proposing the Dartmouth workshop. The proposal, requesting funds from<br>the Rockefeller Foundation was written in August 1955, and is<br>the source of the term artificial intelligence. The term was<br>chosen to nail the flag to the mast, because I (at least) was<br>disappointed at how few of the papers in Automata Studies dealt<br>with making machines behave intelligently. We wanted to focus the<br>attention of the participants.
The original idea of the proposal was that the participants would<br>spend two months at Dartmouth working collectively on AI, and we hoped<br>would make substantial advances.
It didn't work that way for three reasons. First the Rockefeller<br>Foundation only gave us half the meney we asked for. Second, and this<br>is the main reason, the participants all had their own research<br>agendas and weren't much deflected from them. Therefore, the<br>participants came to Dartmouth at varied times and for varying lengths<br>of time.
Two people who might have played important roles at Dartmouth were<br>Alan Turing, who first uderstood that programming computers was the<br>main way to realize AI, and John von Neumann. Turing had died in<br>1954, and by the summer of 1956 von Neumann was already ill from the<br>cancer that killed him early in 1957.
What did happen that summer at Dartmouth?
Newell and Simon, who only came for a few days, were the stars of the<br>show. They presented the logic theory machine and compared its output<br>with protocols from student subjects. The students were not supposed<br>to understand propositional logic but just to manipulate symbol<br>strings according to the rules they were given. They also described<br>representing formulas by list structures and their IPL language.
I thought list structures were a great idea but didn't like the IPL<br>language and immediately thought of using Fortran augmented by list<br>processing primmitive operations coded in machine language.<br>Fortran manuals existed at the time, but an operational Fortran wasn't<br>quite ready.
Alex Bernstein of IBM presented his chess program under construction.<br>My reaction was to invent and recommend to him alpha-beta pruning. He<br>was unconvinced.
My ideas about representing common sense knowledge and reasoning in<br>mathematical logic were still too ill formed for me to present them.<br>Maybe if there had been some logicians at the meeting I'd have hoped<br>for their interest and help. It was another two years before I was<br>ready to present a paper on the subject.
Minsky presented his idea for a plane geometry theorem prover which<br>would avoid much combinatorial explosion by only attempting to proved<br>statements that were true in a diagram. Nat Rochester took this idea<br>back to IBM with him and set Herbert Gelernter, a new IBM hire, to<br>work on it with me as a consultant. Gelernter developed the...