Darwin Among the Machines (18 page)

Von Neumann consulted extensively with the EDVAC group in late 1944 and early 1945 and then, in a virtuoso performance, wrote up a detailed treatment of the engineering principles, logical architecture, and programming language (“order code”) of the proposed computer, submitting the manuscript to the Moore School for review. Herman Goldstine had the incomplete draft (dated 30 June 1945) typed up, with von Neumann listed as sole author, and distributed it widely, with controversial results. On the one hand, the release of the EDVAC report parted the veil of secrecy that had obscured the ENIAC and Colossus projects during the war. The explicit instructions provided in the EDVAC report inspired a flurry of computer building and coding around the world—especially in England, where the Bletchley Park alumni remained handicapped by a prohibition against discussing their own existing work. On the other hand, the attribution of sole authorship to von Neumann embittered Eckert and Mauchly, who left the Moore School to found the Eckert-Mauchly Computer Company,
producer of the BINAC and UNIVAC and ultimately acquired by Sperry-Rand. They believed, with some justification, that von Neumann's report had undermined their interest in future patents by placing the EDVAC design in the public domain. Insult was added to injury by von Neumann's eagerness to propagate the technology as widely and freely as possible, not only in conjunction with the government and academia, but also in cooperation with Eckert and Mauchly's competitors, such as RCA and IBM.

Von Neumann's computer project at the Institute for Advanced Study, launched at the end of 1945, received the bulk of its support not from industry but from the army, the navy, and the Atomic Energy Commission (AEC). Commercial benefits flowed mainly in reverse, with IBM and other organizations partaking freely of the IAS design and IAS-trained personnel. The Institute shied away from industrial contracts but had no qualms about accepting the support of Army Ordnance, the Office of Naval Research, and the AEC. Of some $772,000 of support for the IAS computer project between its inception in 1946 and June 1950, only $82,000 (excluding von Neumann's salary) was contributed by the IAS.
³⁷

Lewis Strauss, J. Robert Oppenheimer, and von Neumann all held influential positions at both the Institute and the AEC. Eventually this surfeit of influence presented a problem, because, as Herman Goldstine, administrative director of the computer project, explained, “when the Atomic Energy Commission up [and] decided one day that it was wrong for the Atomic Energy Commission to engage in research on electronic computers, we had nobody we could go to without all this fear of conflict of interest.”
³⁸
By the time AEC support of the IAS computer project wavered, the IAS design was being replicated widely and a derivative version (first known as the defense calculator) was being developed commercially by IBM. Originally targeted at defense contractors who were building new weapons systems, IBM renamed it the model 701 when they discovered the extent of the demand. Von Neumann was hired as a consultant, officially working thirty days a year for IBM.

Von Neumann's computer bore the paternity of war but, like the jet airplane or the Jeep, it did not remain exclusively a war machine for long. The defense industry employed the brightest minds of the time, individuals who commanded both a clear vision of the future of computers and the resources to bring this future about. “Over the years, the constant and most reliable support of computer science—and of science generally—has been the defense establishment,” concluded Nicholas Metropolis and Gian-Carlo Rota in introducing a
symposium on the history of digital computers at Los Alamos in 1976. “While old men in congresses and parliaments would debate the allocation of a few thousand dollars, farsighted generals and admirals would not hesitate to divert substantial sums to help oddballs in Princeton, Cambridge, and Los Alamos.”
³⁹
The oddballs turned out to be right.

The hydrogen bomb is a three-stage device. Thermonuclear fusion is triggered by a nuclear fission explosion, which is triggered by a high-explosive charge. With no room for trial and error, simulations executed by high-speed computers were as essential to successful bomb building as any of the other ingredients consumed—and transformed—along the way. Perhaps because of this close association from birth with bombs, electronic digital computers acquired an aura of explosiveness that lingers to this day. While computers were being used to catalyze this three-stage process, a series of repercussions was being reflected the other way. From one perspective, computers were testing bombs. From another perspective, bombs were testing computers, unleashing equally powerful results.

For fifty years, the bombs have remained under control. Our worst nightmare has become less of a nightmare as the century draws to a close. Of von Neumann's two creations, it is the computers that exploded, not the bombs.

I
was eight years old, in 1961, when I stumbled on some relics of John von Neumann's electronic computer project left to molder away in an old barn. The barn was itself a relic, predating the establishment of the Institute for Advanced Study amid the Princeton, New Jersey, fields of Olden Farm. Littered with bales of hay, spring-toothed harrows, and other remnants of its working life, the barn now served as auxiliary storage to the Institute's physical plant and, on weekends, as a way station to a small band of boys who hunted frogs and turtles in the swamps and slow-moving streams that bordered the Institute woods. Ancient instincts drew us toward capturing small animals and dismantling machines. Our eyes adjusted to the darkness as a few rays of sunlight perforating the roof traced downward through the dust raised by the pigeons that fluttered away from us overhead. When we stopped talking, absolute silence reigned.

The old barn was a refuge to an extended family of ghosts. Something about abandoned machines—the suspension of life without immediate decay—evokes a mix of fear and hope. When the machine stops, we face whatever it is that separates death from life.

Our fathers were field theorists. At the Institute it was easier to find an expert in celestial mechanics than to find someone who worked on his or her own car. Agricultural implements were as foreign to us as were the mysterious contents of a series of heavy wooden crates piled in the center of the barn, filled with thick wood-and-metal plates of Mediterranean antiquities, the work of one
of the Institute's classical scholars awaiting a second printing that never came. After determining that the plates were the imprints of treasures and not treasures themselves, we scavenged on. Like so many grave robbers before us, we discovered that someone else had gotten to the good stuff first. At one end of the barn was a stockpile of war-surplus electronic equipment that had been selectively cannibalized for vacuum tubes and other vital parts. Partially eviscerated carcasses were distributed like livestock among the abandoned stalls.

We inspected cautiously; then we grew bold and returned with borrowed crescent wrenches and screwdrivers tucked under our belts. First we dismantled relays, making off with small electromagnets that we hooked up to battery power or doorbell transformers at home. Later we discovered microswitches: micro not in size but in the hair-trigger mechanism that shifted their internal state between off and on. Embedded within a maze of wiring and armored in Bakelite, they became the prized trophies of our hunt. Relays, solenoids, and microswitches were thoroughly intertwined. Relays were wired to solenoids wired to microswitches connected to other relays in turn, or sometimes back to the same relays once again. We blindly dissected the fossilized traces of electromechanical logic out of which the age of digital computers first took form. The primitive hardwired architecture, so accessible to our screwdrivers, remained impenetrable to our minds.

The square mile of fields and woodland surrounding the Institute for Advanced Study was cultivated, in lieu of forestry or agriculture, as a sanctuary for ideas. Its founders, in 1930, envisioned their educational utopia as a refuge from the mind-numbing bureaucracy of U.S. universities; they did not imagine the international upheaval from which their enclave would shortly offer an escape. “The Institute was a beacon in the descending darkness,” wrote Director Harry Woolf in 1980, reflecting on the first fifty years, “a gateway to a new life, and for a very few a final place within which to continue to work and transmit to others the style and the techniques of great learning from the other shore.”
²

After the war the Institute became a permanent home to Albert Einstein, Kurt Gödel, John von Neumann, George Kennan, and other scholars equally distinguished if less well known. J. Robert Oppenheimer reigned as director from 1947 to 1966, presiding over what he described as an “intellectual hotel.” He maintained the Institute's lead in mathematical physics while hosting transient scholars as diverse as child psychologist Jean Piaget and poet T. S. Eliot, a visiting member for the fall term of 1948 who listed
The Cocktail Party
(1950) as his only
“publication related to IAS residence.”
³
The Institute woods, bordered by the meandering bends of Stony Brook, offered sanctuary to indigenous wildlife as well, a refuge against the suburban fringe that was metastasizing up and down the eastern seaboard as inexorably as Dutch elm disease, consuming farmland as well as forest and leaving two-car garages in its wake.

The Institute for Advanced Study was conceived by Abraham Flexner (1866–1945), who rose from a Kentucky childhood as the son of an immigrant peddler to become first a schoolmaster and then an influential critic and leading reformer of higher education in the United States. Flexner credited his parents, Esther and Moritz Flexner, with being “shrewd enough to realize that their hold upon their children was strengthened by the fact that they held them with a loose rein.”
⁴
This view formed the guiding principle of Flexner's educational career, even though “to be sure, we shall thus free some harmless cranks.”
⁵
Academic freedom was not to be confused with a lowering of standards for academic work. Flexner emphasized “the impossibility of combining a tender regard for mediocrity with real enthusiasm for learning”
⁶
and secured his reputation as an educator, in 1887, by flunking an entire class.

The Institute appeared as a windfall late in Flexner's life. Flexner was sixty-four years old in 1930 and, as he recalled, “working quietly one day when the telephone rang and I was asked to see two gentlemen who wished to discuss with me the possible uses to which a considerable sum of money might be placed.”
⁷
His visitors represented Louis Bamberger and his sister Caroline (Mrs. Felix) Fuld, retail merchants turned philanthropists after selling the Bamberger department store chain to R. H. Macy & Co., just in advance of the 1929 stock market crash. Flexner persuaded the Bambergers to underwrite not the medical college or local university they had originally intended but an institute that Flexner would later describe as “a paradise for scholars who, like poets and musicians, have won the right to do as they please.”
⁸

This was Flexner's chance to give substance to what he had been preaching about the deficiencies of higher education and research institutions for so many years. “Universities . . . are overorganized,” was his main complaint. The Institute's goal was to avoid “dull and increasingly frequent meetings of committees, groups, or the faculty itself. Once started, this tendency toward organization and formal consultation could never be stopped.”
⁹
The Institute for Advanced Study was incorporated on 20 May 1930, with Flexner as first director, followed by Frank Aydelotte in 1939 and J. Robert Oppenheimer in
1947. Albert Einstein and Oswald Veblen were appointed to the first professorships at the end of 1932, joined by John von Neumann, Hermann Weyl, and James Alexander in 1933. The School of Mathematics opened in 1933, followed by Humanistic Studies and Economics in 1935, Historical Studies in 1948, Natural Sciences in 1966, and Social Science in 1973. Theoretical Biology is rumored to be next.

“The Institute is, from the standpoint of organization, the simplest and least formal thing imaginable,” explained Flexner. “Each school is made up of a permanent group of professors and an annually changing group of members. Each school manages its own affairs as it pleases; within each group each individual disposes of his time and energy as he pleases. . . . The results to the individual and to society are left to take care of themselves.”
¹⁰
For its first decade the Institute had no buildings of its own and was incubated within Princeton University, establishing formal and informal relations between the two otherwise autonomous organizations that have continued to this day. “The mathematicians are guests of the Princeton mathematicians in Fine Hall,” wrote Flexner in 1939 (from his own office among the dentists and lawyers of Nassau Street) and “the humanists are guests of the Princeton humanists in McCormick Hall,” while, true to character (and overlooking the Springdale golf course), “the economists now occupy a suite at the Princeton Inn.”
¹¹