Norbert Wiener
Norbert Wiener (1894–1964) was an American mathematician at the Massachusetts Institute of Technology who founded cybernetics, made foundational contributions to the mathematics of random processes and signal filtering, and became one of the first scientists to warn publicly about the social consequences of automation. He is best known for the 1948 book "Cybernetics: or Control and Communication in the Animal and the Machine" and its popular companion, "The Human Use of Human Beings" (1950), though his technical achievements in harmonic analysis and stochastic processes are what earned him his standing among mathematicians.
Wiener was an exceptional child prodigy whose early life shaped his temperament. Born on 26 November 1894 in Columbia, Missouri, the son of Leo Wiener, a scholar of Slavic languages who taught at Harvard and educated his son intensively at home, he entered Tufts College as a young teenager and received his doctorate from Harvard in 1912, at eighteen, with a dissertation in mathematical logic. He studied afterward in Cambridge and Göttingen under figures including Bertrand Russell, G. H. Hardy and David Hilbert. He later wrote about the pressures of his upbringing in the autobiographies "Ex-Prodigy" (1953) and "I Am a Mathematician" (1956). In 1919 he joined the MIT mathematics department, where he remained for the rest of his career.
His interwar mathematics gave a rigorous footing to randomness in continuous time. In 1923 he constructed the mathematical model of Brownian motion now called the Wiener process, the foundation of stochastic calculus and, much later, of diffusion models, financial mathematics and the noise processes used in machine learning. He also developed generalized harmonic analysis, worked with Raymond Paley on Fourier transforms in the complex domain, and with Eberhard Hopf on an integral equation that bears both their names. These tools prepared him for the engineering problem that redirected his career.
During the Second World War Wiener worked on predicting aircraft trajectories for anti-aircraft fire, with the young engineer Julian Bigelow, and the effort became the seed of both cybernetics and modern signal processing. A gun must aim at where a fast target will be after the shell's flight time, but the observed track is corrupted by noise and by the pilot's evasive maneuvers. Wiener framed this as the statistical problem of extrapolating a stationary random signal and derived the least-squares optimal linear predictor. The device he proposed was never the dominant fire-control system of the war, but the theory was published in a classified wartime report and in 1949 as "Extrapolation, Interpolation, and Smoothing of Stationary Time Series," whose cover earned it the nickname "the Yellow Peril." The Wiener filter and its descendants, the Kalman filter among them, remain staples of estimation, from navigation to speech enhancement and noise reduction. The collaboration with Bigelow and the physiologist Arturo Rosenblueth, and the 1943 paper on behavior, purpose and teleology, set the direction toward the general study of feedback.
After the war Wiener turned from weapons to the human consequences of the machines he had helped imagine. In a 1947 piece in The Atlantic Monthly he announced that he would not provide scientific information to those who might use it for military harm, and he declined to take on military-funded projects. "The Human Use of Human Beings" argued that communication and control are the basis of society and warned that automation would reshape labor. In 1949 he wrote to the labor leader Walter Reuther of the United Auto Workers cautioning about the coming effect of automatic machinery on workers, and he argued that the new machines would compete directly with human labor. His late book "God and Golem, Inc." (1964), which won a posthumous National Book Award, extended these worries to learning machines that might behave in ways their makers did not anticipate. He received the National Medal of Science in 1963 and died on 18 March 1964 in Stockholm.
A common misconception is that Wiener was an enemy of technology, or that he predicted mass unemployment on a fixed schedule. His position was narrower and more modern: machines amplify whatever purposes are fed into them, so the danger lies in specifying goals badly or using the tools to devalue people, and the responsibility lies with the humans and institutions deploying them. His warning about a machine pursuing a purpose that its designer would not really endorse anticipates what is now called the alignment problem. He was also a difficult figure within the scientific community, absentminded in lore and at times in conflict with colleagues, and his relationship to the Macy conference circle cooled in the early 1950s.
Read Norbert Wiener as the mathematician who turned the problem of hitting a moving target into a philosophy of feedback, and who then spent his later years insisting that the machines built on that philosophy must be judged by their effects on human beings.