Simulmatics: Simulating Advertising, Data, Democracy, and War in the 1960s

The History of Computing - Een podcast door Charles Edge

Categorieën:

Dassler shoes was started by Adolf Dassler in 1924 in Germany, after he came home from World War I. His brother Rudolph joined him. They made athletic shoes and developed spikes to go on the bottom of the shoes. By 1936, they convinced Jesse Owens to wear their shoes on the way to his gold medals. Some of the American troops who liked the shoes during World War II helped spread the word. The brothers had a falling out soon after the war was over. Adolph founded Adidas while Rudolph created a rival shoe company called Puma. This was just in time for the advertising industry to convince people that if they bought athletic shoes that they would instantly be, er, athletic. The two companies became a part of an ad-driven identity that persists to this day. One most who buy the products advertised hardly understand themselves. A national identity involves concentric circles of understanding. The larger a nation, the more concentric circles and the harder it is to nail down exactly who has what identity. Part of this is that people spend less time thinking about who they are and more time being told who they should want to be like. Woven into the message of who a person should be is a bunch of products that a person has to buy to become the ideal. That’s called advertising.  James White founded the first modern advertising agency called ‘R. F. White & Son' in Warwick Square, London in 1800. The industry evolved over the next hundred or so years as more plentiful supplies led to competition and so more of a need to advertise goods. Increasingly popular newspapers from better printing presses turned out a great place to advertise. The growth of industrialism meant there were plenty of goods and so competition between those who manufactured or trafficked those goods. The more efficient the machines of industry became, the more the advertising industry helped sell what the world might not yet know it needed. Many of those agencies settled into Madison Avenue in New York as balances of global power shifted and so by the end of World War II, Madison Avenue became a synonym for advertising. Many now-iconic brands were born in this era. Manufacturers and distributors weren’t the only ones to use advertising. People put out ads to find loves in personals and by the 1950s advertising even began to find its way into politics. Iconic politicians could be created.  Dwight D Eisenhower served as the United States president from 1953 to 1961. He oversaw the liberation of Northern Africa in World War II, before he took command to plan the invasion of Normandy on D Day. He was almost universally held as a war hero in the United States. He had not held public office but the ad men of Madison Avenue were able to craft messages that put him into the White House. Messages like “I like Ike.” These were the early days of television and the early days of computers. A UNIVAC was able to predict that Eisenhower would defeat Adlai Stevenson in a landslide election in 1952. The country was not “Madly for Adlai” as his slogan went.  ENIAC had first been used in 1945. MIT Whirlwind was created in 1951, and the age of interactive computing was upon us. Not only could a computer predict who might win an election but new options in data processing allowed for more granular ways to analyze data. A young Senator named John F. Kennedy was heralded as a “new candidate for the 1960s.” Just a few years later Stephenson had lambasted Ike for using advertising, but this new generation was willing to let computers help build a platform - just as the advertisers were starting to use computers to help them figure out the best way to market a product. It turns out that words mattered. At the beginning of that 1960 election, many observed they couldn’t tell much difference between the two candidates: Richard Nixon and John Kennedy. Kennedy’s democrats were still largely factored between those who believed in philosophies dating back to the New Deal and segregationists. Ike presided over the early days of the post-World War II new world order. This new generation, like new generations before and since, was different. They seemed to embrace the new digital era. Someone like JFK wasn’t punching cards and feeding them into a computer, writing algorithms, or out surveying people to collect that data. That was done by a company that was founded in 1959 called Simulmatics. Jill Lepore called them the What If men in her book called If/Then - a great read that goes further into the politics of the day. It’s a fascinating read. The founder of the company was a Madison Avenue ad man named Ed Greenfield. He surrounded himself with a cast of characters that included people from John Hopkins University, MIT, Yale, and IBM.  Ithiel de Sola Pool had studied Nazi and Soviet propaganda during World War II. He picked up on work from Hungarian Frigyes Karinthy and with students ran Monte Carlo simulations on people’s acquaintances to formulate what would later become The Small World Problem or the Six Degrees of Separation, a later inspiration for the social network of the same name and even later, for Facebook. The social sciences had become digital. Political science could then be used to get at the very issues that could separate Kennedy from Nixon. The People Machine as one called it was a computer simulation, thus the name of the company. It would analyze voting behaviors. The previous Democratic candidate Stevenson had long-winded, complex speeches. They analyzed the electorate and found that “I Like Ike” resonated with more people. It had, after all, been developed by the same ad man who came up with “Melts in your mouth, not in your hands” for M&Ms. They called the project Project Microscope. They recruited some of the best liberal minds in political science and computer science. They split the electorate into 480 groups. A big focus was how to win the African-American vote. Turns out Gallup polls didn’t study that vote because Southern newspapers had blocked doing so. Civil rights, and race relations in general wasn’t unlike a few other issues. There was anti-Catholic, anti-Jew, and anti-a lot. The Republicans were the party of Lincoln and had gotten a lot of votes over the last hundred years for that. But factions within the party had shifted. Loyalties were shifting. Kennedy was a Catholic but many had cautioned he should down-play that issue. The computer predicted civil rights and anti-Catholic bigotry would help him, which became Kennedy’s platform. He stood for what was right but were they his positions or just what the nerds thought? He gained votes at the last minute. Turns out the other disenfranchised groups saw the bigotry against one group as akin to bigotry against their own; just like the computers thought they would. Kennedy became an anti-segregationist, as that would help win the Black vote in some large population centers. It was the most aggressive, or liberal, civil-rights plank the Democrats had ever taken up.  Civil rights are human rights. Catholic rights are as well. Kennedy offered the role of Vice President to Lyndon B Johnson, the Senate Majority Leader and was nominated to the Democratic candidate. Project Microscope from Simulmatics was hired in part to shore up Jewish and African-American votes. They said Kennedy should turn the fact that he was a Catholic into a strength. Use the fact he was Catholic to give up a few votes here and there in the South but pick up other votes. He also took the Simulmatics information as it came out of the IBM 704 mainframe to shore up his stance on other issues. That confidence helped him out-perform Nixon in televised debates. They used teletypes and even had the kids rooms converted into temporary data rooms. CBS predicted Nixon would win. Less than an hour later they predicted Kennedy would win. Kennedy won the popular vote by .1 percent of the country even after two recounts. The Black vote hat turned out big for Kennedy. News leaked about the work Simulmatics had done for Kennedy. Some knew that IBM had helped Hitler track Jews as has been written about in the book IBM and the Holocaust by Edwin Black. Others still had issues with advertising in campaigns and couldn’t fathom computers. Despite Stalin’s disgust for computers some compared the use of computers to Stalinistic propaganda. Yet it worked - even if in retrospect the findings were all things we could all take for granted. They weren’t yet. The Kennedy campaign at first denied the “use of an electronic brain and yet their reports live on in the Kennedy Library. A movement against the use of the computer seemed to die after Kennedy was assassinated.  Books of fiction persisted, like The 480 from Eugene Burdick, which got its title from the number of groups Simulmatics used. The company went on to experiment with every potential market their computer simulation could be used in. The most obvious was the advertising industry. But many of those companies went on to buy their own computers. They already had what many now know is the most important aspect of any data analytics project: the data. Sometimes they had decades of buying data - and could start over on more modern computers. They worked with the Times to analyze election results in 1962, to try and catch newspapers up with television. The project was a failure and newspapers leaned into more commentary and longer-term analysis to remain a relevant supplier of news in a world of real-time television. They applied their brand of statistics to help simulate the economy of Venezuela in a project called Project Camelot, which LBJ later shot down.  Their most profitable venture became working with the defense department to do research in Vietnam. They collected data, analyzed data, punched data into cards, and fed it into computers. Pool was unabashedly pro-US and it’s arguable that they saw what they wanted to see. So did the war planners in the pentagon, who followed Robert McNamara. McNamara had been one of the Quiz Kids who turned around the Ford Motor Company with a new brand of data-driven management to analyze trends in the car industry, shore up supply chains, and out-innovate the competition. He became the first president of the company who wasn’t a Ford. His family had moved to the US from Ireland to flee the Great Irish Famine. Not many generations later he got an MBA from Harvard before he became a captain in the United States Army Air Forces during World War II primarily as an analyst. Henry Ford the second hired his whole group to help with the company.  As many in politics and the military learn, companies and nations are very different. They did well at first, reducing the emphasis on big nuclear first strike capabilities and developing other military capabilities. One of those was how to deal with guerrilla warfare and counterinsurgencies. That became critical in Vietnam, a war between the communist North Vietnamese and the South Vietnamese. The North was backed by North Korea, China, and the Soviet Union, the South backed by the United States, South Korea, Australia. Others got involved but those were the main parties. We can think of McNamara’s use of computers to provide just in time provisioning of armed forces and move spending to where it could be most impactful, which slashed over $10 billion in military spending. As the Vietnam war intensified, statistically the number of troops killed by Americans vs American casualties made it look computationally like the was was being won. In hindsight we know it was not.  Under McNamara, ARPA hired Simulmatics to study the situation on the ground. They would merge computers, information warfare, psychological warfare, and social sciences. The Vietnamese that they interviewed didn’t always tell them the truth. After all, maybe they were CIA agents. Many of the studies lacked true scholars as the war was unpopular back home. People who collected data weren’t always skilled at the job. They spoke primarily with those they didn’t get shot at as much while going to see. In general, the algorithms might have worked or might not have worked - but they had bad data. Yet Simulmatics sent reports that the operations were going well to McNamara. Many in the military would remember this as real capabilities at cyber warfare and information warfare were developed in the following decades. Back home, Simulmatics also became increasingly tied up in things Kennedy might have arguably fought against. There were riots, civil rights protests, and Simulatics took contracts to simulate racial riots. Some felt they could riot or go die in in the jungles of Vietnam. The era of predictive policing had begun as the hope of the early 1960s turned into the apathy of the late 1960s. Martin Luther King Jr spoke out again riot prediction, yet Simulmatics pushed on. Whether their insights were effective in many of the situations, just like in Vietnam - was dubious. They helped usher in the era of Surveillance capitalism, in a way. But the arrival of computers in ad agencies meant that if they hadn’t of, someone else would have.  People didn’t take kindly to being poked, prodded, and analyzed intellectually. Automation took jobs, which Kennedy had addressed in rhetoric if not in action. The war was deeply unpopular as American soldiers came home from a far off land in caskets. The link between Simulmatics and academia was known. Students protested against them and claimed they were war criminals. The psychological warfare abroad, being on the wrong side of history at home with the race riots, and the disintegrating military-industrial-university complex didn’t help. There were technical issues. The technology had changed away from languages like FORTRAN. Further, the number of data points required and how they were processed required what we now call “Big Data” and “machine learning.” Those technologies showed promise early but more mathematics needed to be developed to fully weaponize the surveillance everything. More code and libraries needed to be developed to crunch the large amounts of statistics. More work needed to be done to get better data and process it. The computerization of the social sciences was just beginning and while people like Pool predicted the societal impacts we could expect, people at ARPA doubted the results and the company they created could not be saved as all these factors converged to put them into bankruptcy in 1970.  Their ideas and research lived on. Pool and others published some of their findings. Books opened the minds to the good and bad of what technology could do. The Southern politicians, or Dixiecrats, fell apart. Nixon embraced a new brand of conservatism as he lost the race to be the Governor of California to Pat Brown in 1962. There were charges of voter fraud from the 1960 election. The Mansfeld Amendment restricted military funding of basic research in 1969 and went into effect in 1970. Ike had warned of the growing links between universities as the creators of weapons of war like what Simulmatics signified and the amendment helped pull back funding for such exploits. As Lepore points out in her book, mid-century liberalism was dead. Nixon tapped into the silent majority who countered the counterculture of the 1960s. Crime rose and the conservatives became the party of law and order. He opened up relations with China, spun down the Vietnam war, negotiated with the Soviet leader Brezhnev to warm relations, and rolled back Johnson’s attempts at what had been called The Great Society to get inflation back in check. Under him the incarceration rate in the United States exploded. His presidency ended with Watergate and under Ford, Carter, Reagan, and Bush, the personal computer became prolific and the internet, once an ARPA project began to take shape. They all used computers to find and weigh issues, thaw the Cold War, and build a new digitally-driven world order. The Clinton years saw an acceleration of the Internet and by the early 2000s companies like PayPal were on the rise. One of their founders was Peter Thiel. Peter Thiel founded Palantir in 2003 then invested in companies like Facebook with his PayPal money. Palantir received backing from In-Q-Tel “World-class, cutting-edge technologies for National Security”. In-Q-Tel was founded in 1999 as the global technological evolution began to explode. While the governments of the world had helped build the internet, it wasn’t long before they realized it gave an asymmetrical advantage to newcomers. The more widely available the internet, the more far reaching attacks could go, the more subversive economic warfare could be. Governmental agencies like the United States Central Intelligence Agency (CIA) needed more data and the long promised artificial intelligence technologies to comb through that data. Agencies then got together and launched their own venture capital fund, similar to those in the private sector - one called In-Q-Tel. Palantir has worked to develop software for the US Immigration and Customers Enforcement, or ICE, to investigate criminal activities and allegedly used data obtained from Cambridge Analytica along with Facebook data. The initial aim of the company was to take technology developed for PayPal’s fraud detection and apply it to other areas like terrorism, with help from intelligence agencies. They help fight fraud for nations and have worked with the CIA, NSA, FBI, CDC, and various branches of the United States military on various software projects. Their Gotham project is the culmination of decades of predictive policing work.  There are dozens of other companies like Palantir. Just as Pool’s work on Six Degrees of Separation, social networks made the amount of data that could be harvested all the greater. Companies use that data to sell products. Nations use that data for propaganda. Those who get elected to run nations use that data to find out what they need to say to be allowed to do so. The data is more accurate with every passing year. Few of the ideas are all that new, just better executed. The original sin mostly forgotten, we still have to struggle with the impact and ethical ramifications. Politics has always had a bit of a ruse in a rise to power. Now it’s less about personal observation and more about the observations and analyses that can be gleaned from large troves of data. The issues brought up in books like The 480 are as poignant today as they were in the 1950s.

Visit the podcast's native language site