CHAPTER THIRTY-THREE

IBM’S Watson

Vision for Tomorrow

EVERYBODY KNOWS THAT THOMAS WATSON, SR. (1874–1956), built IBM into a big computer company and was a business leader. But “everybody” is wrong. Thomas Watson, Sr., did not build the IBM we now know; his son, Tom Watson, Jr., did this after he moved into the company’s top management in 1946, only thirty-two years old (though the father stayed on until 1956 as a very active chairman). The father’s IBM was never more than medium-size; as late as 1938 it had only $35 million in sales. Of course, it did not make or sell computers; its mainstays were punch-card machines and time clocks.

Instead of being a business leader, Thomas Watson, Sr., did not attain personal success and recognition until he was well past sixty. Twice in the 1930s he personally was on the verge of bankruptcy. What saved him and spurred IBM sales during the Depression were two New Deal laws: the Social Security Act in 1935 and the Wage-Hours Act of 1937–38. They mandated records of wages paid, hours worked, and overtime earned by employees, in a form in which the employer could not tamper with the records. Overnight they created markets for the tabulating machines and time clocks that Thomas Watson, Sr., had been trying for long years to sell with only moderate success.

In 1939, when I was working in New York as correspondent for a group of British papers, I wanted to write a story on Watson and IBM. I had become interested in the company because of the big pavilion it had put up at the New York World’s Fair and thought a story on so small a frog behaving like a big shot might be amusing. “Forget it,” my editor wrote back. “We are not interested in a story on an unsuccessful company which as far as anyone can tell is never going to amount to much.” And Watson was then already sixty-five years old.

But Thomas Watson, Sr., was something far more important than a successful businessman who built a big company. He was the seer and, very largely, the maker of what we now call postindustrial society and one of the great social innovators in American history. Fifty years ago he had the vision of “data processing” and of “information.” In the midst of the Depression and half broke himself, he financed the research that produced both the theoretical foundations for the computer and the first highly advanced model of it.

Fifty years ago or more he also invented and put into practice what in this country is now known, studied, and imitated as Japanese management—needless to say, he did so without the slightest knowledge of Japan and without owing anything to the then-nonexistent Japanese practices.

In fact, Watson’s problem was that he was far ahead of his time, in his vision as well as in his practices.

I first met Watson in the early 1930s. He had been elected president of the American section of the International Chamber of Commerce—a job nobody else had wanted—and I, then a very young reporter, was sent to interview him. But the interview made absolutely no sense to my editor and never appeared. Of course, Watson delivered himself of the platitudes expected from the Chamber of Commerce, about free trade, about international cooperation, and about “the role of the businessman as an ambassador of peace.” But then he started to talk of what clearly was his real interest—things he called data and information. But what he meant by those terms—it was not at all what anyone else at the time meant by them—he could not explain. Indeed, newspapers usually put him down as a crank.

I doubt that Watson himself understood. He had a vision—he saw. But unlike most seers, Watson acted on his vision.

Some kind of computer would have come without him. There were a good many people at the time working at fast calculating machines, especially under the impetus of World War II, with its need for navigational equipment for fast-moving airplanes, for firing mechanisms for long-range cannon shooting at invisible targets way beyond the horizon, for aerial bombs, and for antiaircraft guns. But without Watson it could have been a very different computer, a “calculating machine” rather than an “information processor.” Without Watson and his vision the computer would have emerged as a “tool” rather than as a “technology.”

Watson did not invent a single piece of hardware. He had no technical education, indeed little formal education altogether, having gone to work as a salesman of sewing machines, pianos, and organs in his native upstate New York when barely eighteen. He also had no technical talent, or much grasp of mathematics or of theory altogether. He was definitely not an inventor in the mold of Edison or his contemporary Charles Kettering. The computers that he sponsored, financed, and pushed for—one finished in 1943, another four years later, in 1947—contributed little by way of hardware and engineering solutions even to IBM’s own early commercial computer in the 1950s.

But Watson saw and understood the computer fifteen years before the term was even coined. He knew right away that it had to be radically different from a high-speed calculator. And he did not rest until his engineers had built the first operational true “computer.” Watson specified very early—in the late 1930s at the latest—what computer people now call the architecture of a computer: capacity to store data; a memory, and random access to it; ability to receive instructions and to change them, to be programmed, and to express logic in a computer language. Watson’s 1947 Selective Sequence Electronic Calculator (SSEC) was much more powerful and far more flexible than any machine then in existence or being planned. Its 12,500 vacuum tubes and 21,400 relays could, for instance, solve partial differential equations. It was, above all, the first—and for several years the only—machine to combine electronic computation with stored programs, with its own computer language, and with the capacity to handle its instructions as data: that is, to change, to correct, and to update them on the basis of new information. These are, however, precisely the features that distinguish a computer from a calculator.

When IBM embarked on designing a computer to be sold in large quantities, it applied hardware largely developed by others—mainly university labs at MIT, Princeton, Rochester, and Pennsylvania. But it was all designed to Watson’s original specifications. This explains why the 650, IBM’s first successful commercial machine, immediately became the industry leader and the industry standard when it was introduced in 1953; why it sold eighteen hundred units during its first five years on the market—twice as many as the best market research had predicted total computer sales worldwide to come to in the entire twentieth century; and why it gave IBM the world leadership in computers the company still holds.

The early technical history of the computer is quite murky. And what part Watson and IBM, or anyone else, played in it is remarkably controversial. But there is no doubt at all that Watson played a key role in the conceptual history of the computer. There are many others who were important, if not central, in the engineering of the computer. But Watson created the computer age.

The story begins in 1933 when IBM, on Watson’s orders, designed a high-speed calculator for Columbia University’s scientists and built it out of standard tabulating-machine parts. Four years later, in 1937, Howard Aiken, a Harvard mathematician, suggested that IBM try to link up several existing machines—designed originally for bookkeeping purposes—to create an even faster calculator for the time-consuming and tedious computations needed by astronomers. The work, Aiken estimated, would take a few months and would cost around $100,000. Instead, the work took six years and the cost was well over half a million dollars. For instead of the highspeed calculator Aiken had in mind, Watson went ahead and produced a computer.

First Watson specified an all-electronic machine; all earlier designs, including IBM’s own machines, were electromechanical with levers and gears doing the switching. Of course, electronic then meant using vacuum tubes—the transistor was well into the future. And the resulting machines were therefore big and clumsy, tended to overheat, and used a great deal of electricity. Still, no one at the time even knew how to use electronic devices as switches—the key, of course, to computer technology. Even the needed theoretical work had yet to be done. For Watson and IBM to decide to desert the well-known and thoroughly familiar electromechanical technology for the totally unexplored electronics was a leap into the dark, but it was the decision that made the computer possible.

Watson’s second specification was equally innovative. His project envisaged what no engineer at the time even dreamed of: a computer memory. Howard Aiken in his research proposal had been concerned with rapid calculation; he wanted a “numbers-cruncher.” His idea was to build something that would be able to do very quickly what traditional adding machines and slide rules did slowly. IBM added to this the capacity to store data. And this meant that IBM’s projected machine would be able to process information. It would have a data bank—a term that, of course, did not exist at the time—and would be able to refer back to it. Thus it could memorize and analyze. It could also—and this is crucial—correct and update its own instructions on the basis of new information.

And this then meant—another one of IBM’s then quite visionary specifications—that IBM’s machine was going to be capable of being programmed, that is, of being used for any information capable of being expressed in a logical notation. All the designs then being worked on were for single-purpose machines. The best known of these, called ENIAC, constructed at the University of Pennsylvania during the World War II years for Army Ordnance and completed in 1946, was, for instance, designed to do high-speed calculations for fast-firing guns. But it could do little more, as it had neither memory nor programming capacity. IBM, too, had specific applications in mind—especially calculations for astronomical tables. But, perhaps because Watson knew firsthand of the data-processing needs of a host of institutions—public libraries, the Census Bureau, banks, insurance companies, and the like—his specifications called for a machine capable of being programmed for all kinds of data. IBM’s machine was a multipurpose machine from the start. This, then, also meant that when IBM finally had a computer that could be manufactured and sold—in the early 1950s—it was both willing and able to supply the users whose demand actually created the computer industry, but of whom no one had even thought in the early stages of computer design: business people who wanted computers for such mundane, unscientific purposes as payroll and inventory. Watson’s insistence on memory and program thus explains in large measure why there is a computer industry.

Because of Watson’s specifications of memory and program, the IBM research project of 1937 helped create computer science. The “analytical machine” that Watson tried to develop needed both computer theory and computer language—again, of course, terms that did not exist then. It needed, of course, first-class engineering, as did all the other calculating machines then designed. But it also created, or at least stimulated, the first computer scientists.

This research produced a prototype in late 1943—called the Automatic Sequence Controlled Calculator (ASCC)—and Howard Aiken did do astronomical calculations on it, in early 1944. But it was far too advanced for the manufacturing technology of its time. In fact, only recently have the materials and the technology become available to manufacture the kind of computer for which Watson organized the research in 1937. Watson then donated the prototype, together with money to run and maintain it, to Harvard and immediately commissioned the work on the next design—the work out of which the SSEC evolved several years later. Its first public demonstration, on January 27, 1948, in New York City—at which it calculated all past, present, and future positions of the moon—ushered in the computer age. Even now, thirty-five years later, the computer that popular cartoons show, with its blinking lights and whirling wheels, is the SSEC as it was displayed at this first public demonstration. And looking at Watson’s specifications in retrospect, this is precisely what must have been in his mind’s eye all along.

As significant as Watson the computer seer is Watson the social innovator. He was even further ahead in his social vision than he was in his vision of data and information and had perhaps even more of an impact—and was equally (or more) misunderstood by his contemporaries.

Watson became a national figure in 1940, when Fortune published a violent, mudslinging attack on him. It insinuated that Watson was the American Führer and tried altogether to make him the symbol of everything repulsive in American capitalism. And it also fixed the popular perception of Thomas Watson as a reactionary and as a paternalist, a perception that still lingers on. But today it is clear that Watson’s real offense was to be far ahead of his time.

Rereading the story in 1983, forty-three years later, I saw right away that what had most irked the Fortune writer in 1940 was Watson’s ban on liquor in IBM’s employee clubs and at IBM parties. He was indeed passionately opposed to alcohol—I once heard that his father or one of his uncles had been an alcoholic. But whatever Watson’s reasons, it is perhaps not such a bad idea to keep work and liquor apart.

Then, of course, there was Watson’s regimentation of his sales force, that is, his demand that IBM’s salesmen wear dark suits and white shirts. But what Watson tried to do, as he had explained to me in my interviews with him a year before the Fortune story, was to instill self-respect in salesmen and respect for them among the public. When Watson began, the salesman was a “drummer,” disreputable and a smelly crook, who spent his evenings in the small town’s whorehouse unless he spent it debauching the farmer’s innocent daughter in the hayloft. Watson, who had been a drummer himself, had deeply suffered from the contempt for his calling and was determined that his associates have self-respect and be respected. “I want my IBM salesmen,” he had said in his 1939 talks with me, “to be people to whom their wives and their children can look up. I don’t want their mothers to feel that they have to apologize for them or have to dissimulate when they are being asked what their son is doing” (and there was a clear implication, I thought then, that he was speaking of his own mother).

Watson’s heinous crime, however, and the one the Fortune writer could not forgive him for, was that he believed in a worker who took responsibility for his job, was proud of it and loved it. He believed in a worker who saw his interests as identical to those of the company. He wanted, above all, a worker who used his own mind and his own experience to improve his own job, the product, the process, and the quality. Watson’s crime—and it was indeed seen as a crime in the late 1930s, when a vulgar Marxism infested even fairly conservative members of the intelligentsia—was to make workers enjoy their work and thus accept the system rather than feel exploited and become revolutionary proletarians.

The way Watson did this was, first, to give his employees what we now call lifetime employment. As in Japan today, there was no contractual obligation on the part of Watson’s IBM not to lay off people. But here was a moral commitment. In the first Depression years, 1931 and 1932, IBM did indeed lay off a few people, but Watson immediately stopped it. His refusal later, in the darkest Depression years of 1933 and 1934, to lay off workers could have bankrupted IBM. And employment security for workers has been IBM practice for fifty years now.

In the mid-1930s Watson abolished “foremen.” They became “managers,” whose job it was to assist workers, to make sure that workers had the tools and the information they needed, and to help them when they found themselves in trouble. Responsibility for the job itself was placed firmly on the work group.

The worker, Watson argued, knew far better than anyone else how to improve productivity and quality. And so around 1935 he invented what we now know as quality circles and credit to the Japanese. The industrial engineer was to be a “resource” to the work group and its “consultant,” rather than the “expert” who dictated to it. Watson also laid down that the individual worker should have the biggest job possible rather than the smallest one, and as far back as 1920 he initiated what we now call job enrichment. (And after the old man’s death his son, in 1958, put into effect his father’s old idea and put all employees on a monthly salary, abolishing both the hourly wage and the distinction between blue-collar and white-collar employees.)

Finally, every employee at IBM had the right to go directly to the company’s chief executive officer, that is, to Watson, to complain, to suggest improvements, and to be heard—and this is still IBM practice.

Much earlier Watson had invented what the Japanese now call continuous learning. He began in the early 1920s with his salespeople, who were called into headquarters again and again for training sessions to enable them to “do better what they are already doing well.” Later Watson extended continuous learning to the blue-collar employees in the plant, and in the 1940s he hired Dwayne Orton, a college president from San Francisco, to be the company’s director of education.

These practices very largely explain IBM’s success. They exorcised the great bugaboo of American business, the employees’ alleged resistance to change. They enabled the company in the 1950s and 1960s to grow as fast as any business had ever grown before—and perhaps faster—without much internal turbulence, without horrendous labor turnover, and without industrial strife. And because Watson’s salesmen had all along been used to learning new things, middle-aged punch-card salesmen without experience in computers or in big-company management could overnight become effective executives in a big high-technology company.

But forty years ago Watson’s policies and practices were very strange indeed. They either made Watson look like a crank, which is what most people believed, or they made him into the sinister force the Fortune writer saw. Today we realize that Watson was only forty or fifty years ahead of his time. Watson was actually one year older than Alfred Sloan. But whereas Sloan in the 1920s created modern management by building General Motors, and with it the modern “big corporation,” Watson ten years later and quite independently created the “plant community” that we know to be the successor to Sloan’s “big business enterprise” of the 1920s. He created in the 1930s the social organization and the work community of the postindustrial society.

The first ones to see this, by the way, were the Japanese. Again and again I have been laughed at in Japan when I talk about Japan’s management embodying Japanese values. “Don’t you realize,” my Japanese friends ask, “that we are simply adapting what IBM has done all along?” And when I ask how come, they always say, “When we started to rebuild Japan in the 1950s, we looked around for the most successful company we could find—it’s IBM, isn’t it?”

Watson also anticipated the multinational. And this, too, no one in 1940 could understand.

Watson very early established foreign branches of IBM, one in France, one in Japan, and so on. IBM actually had little foreign business at the time, and Watson knew perfectly well that he could have handled whatever there was through exporting directly. But he also realized that the IBM he envisaged for the future, the IBM of what we now call data processing and information handling, would have to be a multinational company. And so he created, more than fifty years ago, the structure he knew IBM would have to have and would have to know how to manage were it ever to become the company he dreamed of. He also saw that the central problem of the multinational is to combine autonomy of the local affiliate with unity of direction and purpose of the entire company. This is in large measure the reason that Watson deliberately created what to most people seemed such extreme paternalism—today we would call it the IBM culture. Way back in 1939 he tried to explain it to me. “Foreign affiliates,” he said,

Most other multinationals still struggle with the problem Watson foresaw—and solved—fifty years ago.

Watson was an autocrat, of course. Visionaries usually are. Because visionaries cannot explain to the rest of us what they see, they have to depend on command.

Watson was a generous person but a demanding boss and not one bit permissive. But he demanded the right things: dedication and performance to high standards. He was irascible, vain, opinionated, a publicity hound, and an incorrigible name-dropper. And as he grew older he became increasingly addicted to being grossly flattered. But he was also intensely loyal and quite willing to admit that he had been wrong and to apologize. The people who worked for him feared him—but almost no one left.

Of course, he did not fit into the intellectual climate of the New York of the late 1930s. He was already sixty-five at that time and his roots were in the small rural towns of the 1880s, with their church socials and their service-club luncheons, with communal singing and pledges of allegiance to the flag—all the things that the “smart set” of the time had learned to sneer at as the marks of the “booboisie.” And what was worse and made Watson a menace was that his employees shared Watson’s values, rather than those of the intellectuals, and loved to work for him.

Watson was an extraordinarily complex man who fit into no category. He was among the first American businessmen to use modern design and modern graphics for the company logo and company stationery, products, and offices. But he did so mainly because it was the cheapest way to make his company stand out—and he had no advertising money. To get attention for IBM he used the slogan THINK, which he had invented at NCR. His salesmen used THINK notebooks by the thousands—again, it was the only advertising he could afford. And although he himself didn’t have much of a sense of humor, he made up a good many of the THINK jokes and saw to it that they got wide circulation—again, to get attention for his company. Finally he hit upon the brilliant idea of building a huge pavilion at the 1939 New York World’s Fair as the only way to make millions of visitors from all over the world aware of his then still small and practically unknown company and, as he told me at the time, “at the lowest cost per thousand viewers anyone ever had to pay.”

Watson’s popular image is that of the hidebound conservative, and he lent credence to this view by wearing the stiff Herbert Hoover collar long after it had gone out of fashion. But, almost alone among the businessmen of his time, he was an ardent New Dealer and a staunch supporter of Franklin D. Roosevelt. Indeed, FDR offered him major posts in his administration: first secretary of commerce, and then ambassador to Great Britain (when Watson said no, the job went to Joseph Kennedy instead). And at the very end of his life Watson urged President Eisenhower to return to New Deal principles and was upset over Ike’s insistence on reducing the government’s role in economy and society.

Watson had a most unusual personal history, perhaps the most unusual one in the annals of American business. He started out dirt poor as a teenage drummer—not much more than a peddler. He rose to become the star salesman for the then-new National Cash Register Company of Dayton, Ohio, which was the first business-machine company in history. In about fifteen years he had become the company’s sales manager. But then the company was indicted for antitrust violations, with which, we now know, Watson had little or nothing to do.

The trial was a sensation, for National Cash had been the growth company of the early 1900s. Watson drew a stiff fine and a one-year jail sentence. This conviction was set aside two years later, in 1915, by the Court of Appeals, and the government then dropped the case. But Watson had lost his job; both his career and his reputation were in tatters. It was an experience that marked him for life, an experience that he never forgot.

Already forty, he then had to begin all over again as general manager for a small and until then quite unsuccessful company that owned the punch-card patents. It was not until he was past sixty, in the mid-1930s, that he succeeded in establishing this company, by then renamed IBM, on a firm footing.

Watson—and this, I think, makes the man singularly interesting—was a uniquely American type and one the establishment has never understood throughout our history. He was possessed of a towering intellect but was totally nonintellectual. He belongs to the same type as Abraham Lincoln, who similarly offended the establishment of his day, the cosmopolitan, polished, learned Bostonians who considered themselves so superior to the yokel president in whose cabinet they were forced to serve. He was also similar in many ways to America’s first great novelist, James Fenimore Cooper, who was also sneered at, despised, and pilloried by the intellectuals and the establishment of his time, such as the New England Transcendentalists. They could not make sense of the powerful, tragic, and prophetic vision in which Cooper foretold the end of the American Dream in such books as The Pioneers and The Prairie.

This American type is totally native and owes nothing to Europe—which is one reason that the intellectuals of his time do not know what to do with him. Typically the men of this type have a gift for words—and Watson fully shared it. But they are not men of ideas. They are men of vision. What makes them important is that, like Watson, they act on their vision.

(1983)