Today, people often speak of neural networks as offering a promise of machines that do not need to be programmed. But speaking of those old machines in such a way stands history on its head, since the concept of programming had barely appea…Arthur Samuel’s two questions can help us see why perceptrons worked as well as they did. First, Rosenblatt’s credit-assignment method turned out to be as effective as any such method could be. When the answer is obtained, in effect, by add…
Arthur Samuel
scientist · 2 mentions across 1 reading
In this course
Arthur Samuel was a pioneering computer scientist at IBM who developed early machine learning systems, most notably a checkers-playing program in the 1950s that learned to improve its own play through self-correction rather than explicit programming. The course readings invoke Samuel's conceptual framework—particularly his formulation of the credit-assignment problem—to explain how neural networks like Rosenblatt's perceptron could learn without being programmed in the traditional sense. His work historicizes the false opposition between "programmed" and "learning" machines, showing that the very concept of programming emerged alongside machine learning rather than preceding it.
Mentioned in 1 reading
Appears alongside
People mentioned in the same passages — sorted by co-occurrence weight.