Skip to content

Edmund Berkeley: Giant Brains and the Social Responsibility of Computing

Abstract

Edmund Callis Berkeley (1909–1988) was an insurance actuary who saw the first American computers in the 1940s and spent the rest of his life explaining them to everybody else. He wrote the specification behind Prudential’s early order for a UNIVAC, organized the 1947 meeting that founded the Association for Computing Machinery and served as its first secretary, and in 1949 published Giant Brains, or Machines That Think, the first book about computers for general readers. Its tiny teaching machine, Simon, was built in 1950 and is sometimes called the first personal computer; Berkeley went on to sell computer kits by mail and to publish Computers and Automation, the first computer magazine, which held the first computer art contest in 1963. He left Prudential in 1948 because the company would not let him work against nuclear war, and he spent four decades arguing that computer people were responsible for what their machines were used for.

Harvard and Prudential

Edmund Callis Berkeley was born in New York City in March 1909 (on the 21st in the IEEE Computer Society’s biography, the 20th in the Computer History Museum’s). He took a BA in mathematics and logic at Harvard in 1930 and in 1934 joined the Prudential Insurance Company of America in Newark, where he became an actuary and later its chief research consultant. Prudential ran its policies on punched-card machines, and Berkeley, who had studied logic, went to see the first machines that could do more: among them George Stibitz’s relay Complex Number Calculator at Bell Labs (see George Stibitz and the Complex Number Calculator).

In 1942 he joined the Navy as a mathematician at the Dahlgren proving ground and was assigned to Howard Aiken’s Computation Laboratory at Harvard, where he worked on the Mark II relay computer (see Howard Aiken and the Harvard Mark I). When he returned to Prudential in 1946, he wrote a specification for a computer the company could use for its records. It led in 1947 to Prudential’s contract with Eckert and Mauchly’s new company for one of the first UNIVACs (see UNIVAC and the Eckert-Mauchly Computer Corporation).

The ACM

In the summer of 1947 Berkeley invited a handful of colleagues to discuss forming a society for people who built and used computing machines. With John Curtiss of the National Bureau of Standards he organized the meeting at Columbia University on 15 September 1947 at which 78 people founded the Eastern Association for Computing Machinery. It dropped “Eastern” the next year. Berkeley was its first secretary, until 1953, and kept its membership roster himself (see ACM and The Computer Societies Movement).

The Hazards Project

At Prudential Berkeley was given a research project on the greatest hazards facing the modern world, the kind of thing a life insurer might want to know about. He concluded that the greatest was nuclear war. The company dropped the project and told him not to work against nuclear weapons even in his own time. In 1948 he resigned and set up for himself as Edmund C. Berkeley and Associates, later Berkeley Enterprises, which he ran until his death.

Giant Brains

Giant Brains, or Machines That Think, published in 1949, explained to readers without mathematics what the new machines did and how: and surveyed the machines that existed at the time. The book is generally counted as the first explanation of computers for a general readership.

Simon

At the end of Giant Brains Berkeley described a machine small enough for a reader to understand completely, which he named Simon after the Mother Goose character Simple Simon. In 1950 William Porter, a mechanic, and Robert Jensen and Andrew Vall, electrical engineering graduate students at Columbia, built it from relays for under $300 in parts, finishing in April. Simon handled numbers from 0 to 3, had 32 bits of memory arranged as sixteen two-bit registers, took its program from punched paper tape or five keys, and showed its results on five lamps. It could add, negate, compare and select. Radio-Electronics explained how to build one in thirteen articles from October 1950, and Scientific American described “Simple Simon” in November.

Simon was too small to do any useful work, so calling it the first personal computer stretches the term; it was a teaching machine (see The First Personal Computers). It reached at least one future computer scientist: at twelve, Ivan Sutherland worked on Simon with Berkeley.

Misconception: Simon was the first personal computer

The claim appears on history websites because Simon was small, cheap and meant for one person. With two-bit numbers and no stored program in memory, it could not do practical computation of any kind. See Myths and Misconceptions, where Simon heads the list of contested “first personal computers”.

Kits and Robots

Berkeley Enterprises sold plans, kits and courses by mail. With Oliver Garfield, Berkeley designed and marketed Geniac from 1955, a kit for wiring simple logic circuits, sold with a copy of Claude Shannon’s thesis on relay circuits. After the two fell out in 1958, Berkeley sold an almost identical kit as Brainiac, followed by Tyniac. He also built small robots for exhibitions, among them Squee the Robot Squirrel and two others named Rudy and Franken.

Computers and Automation

In 1951 Berkeley began publishing a roster of organizations in the computing field, which became The Computing Machinery Field in 1952 and Computers and Automation in 1953. It is usually counted as the first magazine about computers. He published it, edited it, and wrote much of it, some of it under the pseudonym “Neil D. MacDonald”. The magazine carried technical papers, product directories and employment listings, and from 1963 it ran an annual computer art contest, the first of its kind; Berkeley is credited with the term “computer art” (see Computer Art and Generative Art). A listing in its June 1965 issue is one of the earliest known printed uses of the phrase “software engineering”; an RCA advertisement in the New York Times used it in 1961.

Berkeley joined the Committee for a SANE Nuclear Policy in 1958 and used the magazine to argue that computer professionals were responsible for how their work was used. In 1970 it went further. Its May issue gave 32 pages to Richard E. Sprague’s analysis of the photographs of the Kennedy assassination, which concluded that the Warren Commission was wrong and that dozens of people had been involved; an October article made the same case for the killing of Robert Kennedy. Introducing it, the magazine said it published such work because the subject was not receiving adequate coverage anywhere else. In 1973 it was renamed Computers and Automation and People, in 1975 Computers and People, and it ceased publication in 1988.

Death

Edmund Berkeley died on 7 March 1988 in Newton, Massachusetts. His papers are at the Charles Babbage Institute in Minneapolis, and the Computer History Museum holds a second collection. The historian Bernadette Longo published his biography, Edmund Berkeley and the Social Responsibility of Computer Professionals, with ACM Books in 2015.

📚 Sources