— Ch. 1 · Origins And Epiphany —
Logic Theorist.
~4 min read · Ch. 1 of 7
In 1954, a RAND scientist named Allen Newell sat in a room while Oliver Selfridge gave a presentation on pattern matching. The air was thick with the hum of early computing equipment and the quiet focus of researchers studying logistics. Newell watched as Selfridge described how simple programmable units could interact to create complex behavior. It happened all at once during that single afternoon. He later recalled having such a sense of clarity that this was a new path he would follow. This moment of scientific epiphany lasted for ten to twelve hours without his normal mode of thought interfering. Simon joined him shortly after, bringing his background in political science and bounded rationality theory. They began discussing whether machines could be taught to think like humans. Their first target was proving mathematical theorems from Whitehead and Russell's Principia Mathematica.
Development And Simulation
The team wrote their initial program onto 3x5 index cards before any computer existed to run it. In January 1956, Herbert Simon gathered his wife, three children, and several graduate students around a table. Each person received one card representing a component of the computer program. They acted out the logic manually to test if the system worked. Nature imitated art imitating nature as they simulated the machine with human bodies. This hand-simulation proved the program could successfully prove theorems just as a talented mathematician might. Eventually Cliff Shaw managed to run the code on actual hardware at RAND's Santa Monica facility. The transition from paper cards to electronic processing marked a pivotal shift in how they approached problem solving.The 1956 AI Conference Debut