Arthur Samuel coined the term “machine learning” in 1959 while working on computer game research.
Samuel was an American pioneer in artificial intelligence and computer gaming. At IBM, he developed a checkers program that could improve its play through experience rather than relying solely on a fixed list of instructions. His work helped demonstrate that computers could use feedback and stored game positions to make better decisions over time.
In a 1959 paper titled “Some Studies in Machine Learning Using the Game of Checkers,” Samuel described methods for enabling a program to learn from playing games. The phrase became a lasting name for approaches in which computer systems identify patterns or improve performance from data or experience.
Samuel’s checkers program is often described as an early machine-learning success, but it was not a modern neural network. It used techniques such as position evaluation, look-ahead search, and self-play. Arthur Samuel should also not be confused with Frank Rosenblatt, who developed the perceptron, an earlier neural-network model.