# Arthur Samuel **Domain:** Artificial Intelligence / Machine Learning / Computing History **Doc Type:** Person Node **Maturity:** Developed **Related:** [[Samuel Checkers]], [[Machine Learning]], [[IBM]], [[Machine Intelligence Continuum]] ## Definition **Arthur Lee Samuel (1901–1990)** was an American computer scientist whose IBM checkers programs became foundational demonstrations of a computer improving its performance through experience. He helped establish the term **machine learning**. ## Contribution Samuel joined IBM in 1949 and developed checkers programs on IBM’s 700-series computers. Because exhaustive search was impossible, the programs evaluated positions, looked ahead selectively, and retained information from prior play. The work showed that useful behavior could be improved through accumulated experience rather than exhaustively specified move-by-move. ## Historical Precision Samuel’s early work was associated with IBM in Poughkeepsie and preceded the 1961 opening of the [[IBM Thomas J. Watson Research Center]]. It belongs to the institutional lineage Yorktown consolidated, but it should not be retrospectively relocated to that building. ## Corpus Function Samuel connects [[Machine Learning]] to [[articles/The Evolutionary Roots of Silicon Valley|The Evolutionary Roots of Silicon Valley]]. Evaluation and retention appear in both learning and evolution, but the mechanisms differ: his program did not biologically reproduce, and “selection” was structured by an engineered game and scoring process. ## Sources / Provenance - IBM, “The games that helped AI evolve”: https://www.ibm.com/history/early-games - IBM, “The history of artificial intelligence”: https://www.ibm.com/think/topics/history-of-artificial-intelligence