The Tinkerers

by Alec Foege

  The term used by RAND for its tinkering procedure was “systems analysis,” coined by a RAND engineer named Ed Paxson in 1947. Paxson, formerly a scientific advisor to the US Armed Forces, had long admired game theory, and wanted to apply its ideas directly to the process of war. Inspired in part by what was known during World War II as operations research (OR), Paxson set out to take America’s defense policy beyond the statistics and hard data that, in his view, weighed it down.

  In operations research, the goals were to figure out how much damage could be inflicted with the resources available to the United States and how efficiently those military plans could be executed to minimize losses. At what altitude should fighter planes fly? How many would be required to achieve a certain goal? Mathematics played a key role in achieving these data-based goals.

  Systems analysis was different from game theory, and distinctly American in its outlook. Instead of relying on existing data to craft solutions to the nation’s current defense goals, systems analysis first reflected upon what the nation hoped to achieve in the future, and how it might craft solutions to those problems. As Alex Abella explains it, in his well-reported book on the RAND Corporation, Soldiers of Reason, “Systems analysis changed the questions and asked: How many enemy factories do we want to destroy? What kind of factories are we talking about and how are they defended? To accomplish our objective, what is the best route? With what kind of plane? What kind of payload?”

  While operations research focused on finding new ways to improve and streamline existing systems, systems analysis took today’s existing knowledge and tinkered with it, until it created the potential problems of tomorrow, as well as an assortment of systems to solve those problems, systems that had yet to be devised.

  Operations research amounted to doing the best with whatever one had. Systems analysis insisted on expanding the palette of options, even if those options didn’t exist yet. In essence, it was a license to dream. Not necessarily to dream only of the destruction of America’s enemies, but rather to dream of a democratic nation that could bend the will of nature to embrace its own desires.

  If that notion sounds a trifle grandiose, the reality was substantially more down to earth. Indeed, in classic RANDian fashion, the dreaming process itself was steeped in mathematics and methodology. Imagined projects were categorized and measured and cost-analyzed as if they were real. If a new fighter plane were part of the equation, systems analysis would determine how fast it would go, how much it would cost to build, how far it would fly, and how much fuel it would use.

  Prompted by the detonation of an atomic bomb exploded by the Soviets in 1949, the US Air Force sought to fashion a preemptive attack on the Communist powerhouse. It assigned Ed Paxson to create a suitable bomber for the plan.

  Paxson’s approach was both creative and disheartening, though perhaps not surprising given the grim goal of the task. Set on developing a “science of war” out of whole cloth, he began by compiling all manner of detail about the aerial bombing capabilities of both sides in the projected conflict. The mountain of resulting data was so large that RAND developed its own early computer to manage the results.

  To top it off, Paxson created a combat simulator in RAND’s basement to allow pilots from the air force and navy to practice their craft against films of real war footage. And Paxson’s resulting report, “Comparison of Airplane Systems for Strategic Bombing,” produced in 1950, ultimately betrayed the tinkering roots of systems analysis.

  Make no mistake: RAND’s approach to systems analysis had a fatal flaw and, as a result, took American tinkering down a dark hole. RAND’s so-called rational approach to national defense helped shape the United States’ policy of counterforce in the 1950s (stockpiling nuclear warheads) and its disastrous response to the Vietnam conflict in the 1960s.

  And while RAND liked to paint its peace-keeping proposals as “realistic,” the truth was they were inordinately pessimistic. By developing worst-case scenarios regarding world politics, systems analysis seemed predisposed to recommending only extreme, apocalyptic solutions to achieving peace.

  President Dwight Eisenhower, a Republican, had warned in his January 1961 farewell address of the “unwarranted influence” of the “military-industrial complex.” Meanwhile, Democrat John F. Kennedy had accessed information compiled by RAND researchers to help fuel his campaign against Richard Nixon during the 1960 presidential election. During the campaign, Kennedy regularly cited America’s defense deficiencies as highlighted in 1957’s Gaither Report, named after its originator, RAND chairman H. Rowan Gaither. What became known as the “missile gap,” the belief that the Soviets were building nuclear missiles at a far faster rate that the Americans, was a valuable tool in the Democrats’ quest to defeat Nixon.

  Never mind that an investigation initiated by Eisenhower a few years earlier revealed that the Gaither Committee had based its report on faulty data. A freshly inaugurated Kennedy was sold on RAND’s rational approach and its groundings in the intellectual elite.

  Once Kennedy was in office, he appointed Robert McNamara his new secretary of defense. While McNamara was not a RAND alumnus, his cool, numbers-based approach to problem solving—honed in his former position as Ford Motor Company’s CEO—was in perfect alignment with the RANDian worldview. Indeed, during World War II, he had performed statistical analysis for General Curtis LeMay, later one of RAND’s founders, that resulted in the American firebombing of Japanese cities. Once ensconced in the Kennedy cabinet, McNamara hired Charles Hitch, head of RAND’s economic division and author of The Economics of Defense Spending in the Nuclear Age, as his deputy.

  McNamara’s Whiz Kids, as the preppy defense secretary and his team became known, quickly began to reorganize the Defense Department to reflect what they saw as a new age of warfare. The air force sustained drastic cuts while the navy’s Polaris submarine program received new focus and resources. Decisions were made with new thriftiness, efficiency, and flexibility. Meanwhile, the Whiz Kids developed an insouciant confidence that offended an entrenched military. Then came the disastrous Bay of Pigs invasion in April 1961, which was planned under Eisenhower but okayed by McNamara in the early days of the Kennedy administration. After CIA-TRAINED Cubans failed to overthrow the Castro government, McNamara concluded that “the government should never start anything unless it could be finished, or the government was willing to face the consequences of failure.”

  What emerged was the strategy of counterinsurgency, a secretive web of tactics that included the establishment of the Green Berets and paramilitary forces tucked in areas of Latin America and Asia.

  After narrowly diffusing the Cuban Missile Crisis in October 1962 through pure luck—McNamara’s suggestion to defuse US missiles in Turkey before invading Cuba instead spurred a behind-the-scenes withdrawal of Soviet missiles from Cuba—the young and inexperienced defense secretary ramped up the implementation of systems analysis as the primary tool of warfare. “Every quantitative measurement we have show we are winning this war,” said McNamara after visiting South Vietnam for the first time in April 1962. According to his statistics, the war would be over in three to four years.

  By the time he was terminated by President Johnson in November 1967, McNamara was a broken, disillusioned man. Haunted by the mounting American deaths in Vietnam, he became convinced that the war was futile, though he would not reveal his personal thoughts until decades later. McNamara’s rational, intellectual, and, dare one say, creative approach to the Vietnam War did little to account for the bloody human toll.

  Decades later, some of these same pessimistic ideas about reality would come to be grouped under the umbrella term neoconservatism, which had its roots in RAND’s approach to national defense. Under President George W. Bush these ideas would reach the height of their influence, as cabinet members Paul Wolfowitz and Richard Perle, two protégés of Albert Wohlstetter, the influential RAND analyst in the 1950s who later taught at the University of Chicago, played a large role in the
Bush administration’s decision to invade Iraq in 2003. And both hardliners helped transform US defense policy from a largely bureaucratic endeavor to a personalized one that reflected the crypto-intellectualized riffs of a few individuals who began experimenting, or tinkering, with their wildest foreign policy theories in real time.

  In place of the free-range thought processes that had always made American innovation so powerful and impactful, RAND had substituted a kind of cold, amoral stubbornness. By deciding to be collectively transformative, RAND’s analysts succeeded only in being collectively transgresssional. In its dyspeptic approach, RAND boiled the humanity out of one of the most innate of human instincts—vision—and muddled the true, wild-man contributions of the American individualist unbridled.

  It would take decades for the idea of team-coordinated innovation to regain zeitgeist status. In its new iteration, new attention would be placed on the value of the lone tinkerer. And Nathan Myhrvold, a product of that idea factory known as Microsoft, would come up with one of the more novel approaches.

  CHAPTER 5

  MYHRVOLD’S MAGIC TINKERING FACTORY

  THE NOTION OF STANDARDIZING or commercializing the process of tinkering is not a new one. After all, Edison had something like that in mind from the time he built his first lab. But the exigencies of the modern world have created a whole new notion of what tinkering can do to help fuel our economy. Venture capital firms, investment operations that fund start-ups, focus on existing companies. This is something one step removed from the direct sponsorship of tinkering.

  In 2000, Nathan Myhrvold sought to fill this gap with the launching of Intellectual Ventures, which he calls the world’s first “invention capital” firm. Myhrvold may be best known as Microsoft’s former chief technology officer and as a phenomenally rich man who has pursued a variety of interests to their logical extremes.

  Intellectual Ventures is Myhrvold’s effort to bridge the gap between inventive tinkering and the realization of those ideas. The firm, which Myhrvold cofounded, he said, “is about investing in inventions. So we either do our own tinkering or we invest in other people’s tinkering. Either way, we are all about the process of invention.”

  The way Intellectual Ventures works, according to Myhrvold, sounds a little like a conventional venture capital firm but with a major twist. The company raises funds from large corporations interested in investing in innovation. Over the past decade, it has amassed a pot of more than $5 billion from a group of big names in the technology industry, including Microsoft, Intel, Apple, SAP, Nvidia, eBay, and Sony, along with some investment firms, such as Charles River Ventures. But instead of using those funds to invest in promising start-ups, it hired a lot of smart people, many of them biotechnologists, physicists, and engineers, to come up with new ideas for commercial products and patent those ideas. Then Intellectual Ventures would market and license the patents to interested corporations. With eight hundred employees and the seventh-largest portfolio of patents in the world, the company certainly wields power.

  The big question is whether it represents the future of tinkering. Myhrvold views his firm as a defender of individual inventors against major corporate interests, even as some critics have suggested he is holding corporations for ransom.

  Speaking with Myhrvold about Intellectual Ventures, one quickly gets the feeling it is the next step in a lifetime of tinkering that has been proven out over and over again in what can only be described as a remarkably unusual and fortunate life.

  Myhrvold was born in 1959 and grew up in Santa Monica, California. Raised by his mother, a schoolteacher, he graduated from high school at fourteen and later attended UCLA, where he received a bachelor’s degree in mathematics and a master’s in geophysics and space physics. By twenty-three, he had earned another master’s in mathematical economics, and a PhD in theoretical physics, both from Princeton.

  As brilliant as he appeared, Myhrvold seemed like an unlikely hire for Microsoft. He was, by most measures, a hyperacademic type not at all interested in less lofty notions such as commerce. The title of his dissertation at Princeton was “Vistas in a Curved Space-Time Quantum Field Theory,” a dead giveaway. “When I was writing my thesis at Princeton in theoretical physics, I started using one of the early computers—this was one even before the IBM PC,” Myhrvold told me.

  He used an early word-processing software called Magic Wand designed for a microcomputer, a predecessor to the personal computer. Then, in the early 1980s, as more powerful PCs, like the Commodore 64 and the NEC PC-98, were beginning to come out, he began writing scientific software designed to calculate complex mathematical problems as well as to create visual models of mathematical concepts. What he was attempting was similar to what later came out as Wolfram’s Mathematica program.

  Myhrvold and a group of friends from Princeton got involved enough in computer programming that he took a leave of absence after his first year at Cambridge University, where he had begun a postdoctoral fellowship in 1984 under the guidance of Stephen Hawking, to return to the United States. Before Myhrvold and his pals designed their mathematical software, however, the group decided they needed to build operating system extensions to facilitate the integration of their software into existing computer frameworks such as DOS.

  Those operating systems extensions formed the basis of the earliest Windows system. When the friends started a company to market their operating system, Myhrvold was made chief executive of the resulting company, Dynamical Systems Research, as his leave of absence from Cambridge became longer and longer. Nonetheless, motivated by solidarity with his friends, he moved to the Berkeley, California, area in the next year and ran the company, which sought to produce a software product called Mondrian, which provided a multitasking environment for DOS. In 1986, Microsoft, then eleven years old, saw Mondrian as a missing key element in its own Windows system and bought the company and the technology.

  Myhrvold was offered a job by Bill Gates, who identified him as a superior thinker he wanted on his side, and Myhrvold finally gave up his spot at Cambridge for good. For his first four years at Microsoft, Myhrvold would report to Steve Ballmer in the operating systems division. Over the next decade, Myhrvold rose within Microsoft to become the chief technology officer. All this, despite the fact that he had no formal computer or engineering training.

  Despite his success, Myhrvold has always viewed himself as an outsider. That perspective, no doubt acquired as an unpopular, nerdy kid, often has freed him to pursue what interests him as opposed to what he was expected to be interested in, and take any of those interests to its logical extreme.

  “I started tinkering when I was a kid,” Myhrvold said. “I took lots of things apart and then put them back together again—of course, the problem was always having extra parts at the end and wondering, Where do those things go?”

  His journey into the world of molecular gastronomy, otherwise known as “modernist cooking,” is a good example of this. Where classical cuisine is concerned primarily with taste and presentation, molecular gastronomy delves into the chemical and physical changes that ingredients go through during the cooking process, seeking to elevate those elements to an equal level with the more traditional concerns.

  An amateur cook since childhood, Myhrvold attended the École de Cuisine La Varenne in Burgundy, France, in the early 1990s, while still employed by Microsoft. There, he became entranced by sous vide cooking, which involves sealing food in airtight plastic packages and immersing it in a warm bath of water for extended periods at extremely precise temperatures. The technique had been around since at least the mid-1970s, but regained popularity among top chefs some two decades later because it allowed them inordinate control over the temperature at which meat or fish was cooked, meaning their dishes would never be overcooked or too rare.

  Myhrvold retired from Microsoft in 2000, at age forty, and within five years he was posting his observations about sous vide cooking on an online bulletin board called eGullet.org. He was drawn
to cooking as another opportunity for innovation. He soon discovered there was no book published in English on the topic of sous vide; so he decided to write one.

  The result evolved into Modernist Cuisine, an idiosyncratic cookbook published in 2011 based on Myhrvold’s experimentation at the Cooking Lab, an offshoot of Intellectual Ventures. Retailing for $625 a copy, the twenty-four-hundred-page cooking compendium includes recipes for carbonated fruit, watermelon that looks like meat, and a macaroni and cheese that includes wheat beer, sodium nitrate, and a gelatin made from red seaweed.

  Deciding to approach cooking from a scientific perspective rather than a culinary one, Myhrvold invested in millions of dollars of highly technical equipment, including an autoclave—designed to sterilize medical equipment with high-pressure steam—a rotary evaporator, and a hundred-ton hydraulic press. In 2007, he established Cooking Lab and hired a team of cooks, writers, photographers, editors, and designers to carry out his off-the-wall food-science tinkering experiments.

  While the resulting six volumes, which include copious photos taken by Myhrvold and unorthodox titles such as Techniques and Equipment and Plated Dish Recipes, are nearly comprehensive in their exploration of cooking’s cutting edge, they offer little for the amateur chef to replicate. Rather they stand as a tinkering achievement, the extreme example of what an individual can achieve in a specific discipline without taking the most direct route to mastering it.

  At his best, Myhrvold is a champion of tinkering for tinkering’s sake. By ignoring the market potential of his culinary experiments and the resulting encyclopedic tome, he freed himself to follow his passion for understanding molecular gastronomy to its limit.

  One of his first tinkering projects was assembling an electronic discharge machining system, known as an EDM tool, used to shape materials with sparks generated by a pair of electrodes. In those days, magazines such as Popular Science and Popular Mechanics featured plans for things you could build that were interesting—and maybe even dangerous. “You could build some pretty wild stuff,” Myhrvold said. The EDM tool was one of those projects. The idea was to use extremely high-voltage electrical current to generate sparks to etch and cut metal. When Myhrvold made his, he was around nine years old.