The Undoing Project

by Michael Lewis

  After Michigan, Danny gave talks about the undoing project and neglected to mention Amos. He’d never done that sort of thing before. For a decade, they had had a hard-and-fast rule against inviting others anywhere near areas of mutual interest. At the end of 1979, or perhaps in early 1980, Danny began to talk to a young assistant professor at UBC named Dale Miller, sharing his ideas about the way people compared reality to its alternatives. When Miller asked about Amos, Danny said that they were no longer working together. “He was in Amos’s shadow and he was very worried about that, I think,” said Miller. It wasn’t long before Danny and Miller were working on a paper together that might just as well have been called “The Undoing Project.” “I thought that they had agreed to see other people,” said Miller. “And he was insistent that his days of collaborating with Amos were over. I remember a lot of fraught conversations. At some point he said to be gentle with him, because this was his first relationship after Amos.”

  * * *

  If the Katz-Newcomb Lecture meant less to Amos than to Danny, it was because Amos’s life was now a sprint from one Katz-Newcomb Lecture to the next. He reminded at least one of his new graduate students at Stanford of a stand-up comic, traveling the world and working small nightclubs to test his material. “He thought by talking,” recalled his wife, Barbara. “You could hear him in the shower. You could hear him talking to himself. Through the door.” His children grew used to hearing their father alone in a room, talking. “It was a little bit like an insane person talking to himself,” said his son Tal. They’d see him coming home in his brown Honda, stopping and starting in the street in front of their house and talking. “He’d be going three miles an hour, then all of a sudden he’d gun it,” said his daughter, Dona. “He’d worked out the idea.”

  In the weeks leading up to the Katz-Newcomb Lecture, in early April 1979, Amos was busy talking to the Soviet Union. He’d joined a delegation of ten prominent Western psychologists on a bizarre intellectual diplomatic mission. Soviet psychologists were then trying to persuade their government to admit mathematical psychology into the Russian Academy of Sciences and had asked their American counterparts for support. Two distinguished mathematical psychologists, William Estes and Duncan Luce, had taken it upon themselves to help them. The older guys made a short list of America’s leading mathematical psychologists. Most of them were ancient. Amos counted as one of the younger guys, along with his Stanford colleague Brian Wandell. “The older guys had this idea that we were going to rescue the image of psychology in the Soviet Union,” recalled Wandell. “Psychology flew in the face of Marxism. It was on the list of things that didn’t need to exist.”

  It took about a day to realize why Marxism might feel that way. These particular Soviet psychologists were charlatans. “We were thinking there really were going to be scientists on the Soviet end,” said Wandell. “There weren’t.” The Soviets and the Americans took turns giving presentations. An American would give a learned talk about decision theory. His Soviet counterpart would rise and offer a talk that sounded completely insane—one guy spent his allotted time on his theory about how the brain waves caused by beer canceled the brain waves caused by vodka. “We’d get up and give a paper, and you know, it was okay,” said Wandell. “Then some Russian guy would get up and talk and we’d say, ‘Wow, that was weird.’ One was about how the meaning of life could be put into a formula and the formula might have some variable labeled E in it.”

  With one exception, the Russians knew nothing about decision theory, and didn’t even seem particularly interested in the subject. “There was one guy,” said Wandell, “who gave this great talk, at least compared to the others.” That guy turned out to be a KGB agent, whose training in psychology consisted of the talk he had given. “The way we discovered he was a KGB guy was that he showed up later at a physics conference and gave a great talk there, too,” said Wandell. “That was the only guy Amos liked.”

  They stayed in a hotel where the toilets didn’t flush and the heat didn’t work. Their rooms were bugged, and everywhere they went they were followed by guards. “People were pretty freaked out the first day or two,” said Wandell. “We were plainly in over our heads.” Amos found the whole thing hysterical. “They put a focus on Amos, probably because he was Israeli,” said Wandell. “In typical Amos fashion, he was walking around Red Square, and gives me this look that says, ‘C’mon, let’s lose ’em!’ Then he just kind of took off, with the guards chasing after him.” When they finally caught up to him—hiding in a department store—the Soviets were furious. “They gave us all a stern talking-to,” said Wandell.

  Amos spent at least some of his time in his bugged and heatless hotel room adding to a file that he’d labeled “The Undoing Project.” The file in the end came to forty or so pages of handwritten notes. Between the lines, you can hear the polite throat clearing of a diamond cutter waiting for his rocks. Amos clearly had hopes of turning Danny’s ideas into a full-blown theory. Danny didn’t know that, or that Amos was busy dreaming up his own vignettes:

  David P was killed in a plane crash. Which of the following is easier to imagine:

  —that the plane did not crash

  —that David P. took another plane

  Instead of replying to Danny’s long letter, Amos made notes to himself, trying to order the stuff spilling out of Danny. “The present world is often surprising, i.e., less plausible than some of its alternatives,” he wrote. “We can order possible worlds by i) initial plausibility and ii) similarity to the present world.” He followed this a few days later with eight dense pages in which he attempted to create a logical, internally consistent theory of the imagination. “He loved these ideas,” said Barbara. “It’s something very basic about decision making that fascinated him. It’s the choice you don’t take.” He groped for titles, so that he might know what he was writing about. In his earliest notes in the file, he scribbled the phrase “the undoing heuristic” and gave the new theory the name “Possibility Theory.” He then changed it to “Scenario Theory,” and then again to “The Theory of Alternative States.” In the last notes he made on the subject, he called it “Shadow Theory.” “The major point of shadow theory,” Amos wrote to himself, “is that the context of alternatives or the possibility set determines our expectations, our interpretations, our recollection and our attribution of reality, as well as the affective states which it induces.” Toward the end of his thinking on the subject, he summed up a lot in a single sentence: “Reality is a cloud of possibility, not a point.”

  It wasn’t that Amos had no interest in Danny’s thoughts. It was that they were no longer talking in the same room, with the door closed. The conversation that he and Danny were meant to be having together, each was more or less having alone. Because of the new distance between them, each was far more aware where the ideas had come from. “We know who had the idea, because of the physical separation and because the idea is in a letter,” Amos would complain to Miles Shore. “Before, we would have picked up the phone at the beginning of an idea. Now you develop an idea and you become committed to them, and they become more personal and you remember you had them. Initially we never had that.”

  Committed to his new idea, Danny had taken it back rather than let Amos take it apart and remake it into something more like his own. Amos continued to fly to Vancouver every other week, but there was a new tension between them. Amos clearly wished to believe that they might collaborate as they had before. Danny did not. He’d anticipated his own envy and built it into a decision about Amos.

  * * *

  * That strange fact comes from an excellent article on the subject of pilot illusions by Tom LeCompte in the Smithsonian’s Air & Space magazine.

  12

  THIS CLOUD OF POSSIBILITY

  Amos was in Israel on a visit in 1984 when he received the phone call telling him that he’d been given a MacArthur “genius” grant. The award came with two hundred fifty thousand dollars, pl
us an extra fifty thousand dollars for research, a fancy health care plan, and a press release celebrating Amos as one of the thinkers who had exhibited “extraordinary originality and dedication in their creative pursuits and a marked capacity for self-direction.” The only work of Amos’s cited in the press release was the work he’d done with Danny. It didn’t mention Danny.

  Amos disliked prizes. He thought that they exaggerated the differences between people, did more harm than good, and created more misery than joy, as for every winner there were many others who deserved to win, or felt they did. The MacArthur became a case in point. “He wasn’t grateful for that prize,” said his friend Maya Bar-Hillel, who was with Amos in Jerusalem shortly after the prize was announced. “He was pissed. He said, ‘What are these people thinking? How can they give a prize to just one of a winning pair? Do they not realize they are dealing the collaboration a death blow?’” Amos didn’t like prizes but he kept on getting them anyway. Before the MacArthur “genius” grant, he had been admitted into the American Academy of Arts and Sciences. Soon after the MacArthur, he received a Guggenheim Fellowship and an invitation to join the National Academy of Sciences. That last honor was seldom bestowed on scientists who weren’t U.S. citizens—and it wasn’t bestowed upon Danny. There would follow honorary degrees from Yale and the University of Chicago, among others. But the MacArthur was the prize Amos would dwell upon as an example of the damage caused by prizes. “He thought it was myopic beyond forgiveness,” said Bar-Hillel. “It was genuine agony. He wasn’t putting on a show for me.”

  Along with the prizes came a steady drizzle of books and articles praising Amos for the work he had done with Danny, as if he had done it alone. When others spoke of their joint work, they put Danny’s name second, if they mentioned it at all: Tversky and Kahneman. “You are very generous in giving me credit for articulating the relationship between representativeness and psychoanalysis,” Amos wrote to a fellow psychologist who had sent Amos his new journal article. “These ideas, however, were developed in discussions with Danny so you should mention both our names or (if that appears awkward) omit mine.” An author of a book credited Amos with noticing the illusory sense of effectiveness felt by Israeli Air Force flight instructors after they’d criticized a pilot. “I am somewhat uncomfortable with the label the ‘Tversky effect,’” Amos wrote to the author. “This work has been done in collaboration with my long-time friend and colleague, Daniel Kahneman, so I should not be singled out. In fact, Daniel Kahneman was the one who observed the effect of pilots’ training, so if this phenomenon is to be named after a person it should be called the ‘Kahneman effect.’”

  The American view of his collaboration with Danny mystified Amos. “People saw Amos as the brilliant one and Danny as the careful one,” said Amos’s friend and Stanford colleague Persi Diaconis. “And Amos would say: ‘It’s exactly the opposite!’”

  Amos’s graduate students at Stanford gave him a nickname: Famous Amos. “You knew that everyone knew him, and you knew everyone wanted to hang out with him,” said Brown University professor of psychologist Steven Sloman, who studied with Amos in the late 1980s. The maddening thing is that Amos seemed almost indifferent to the attention. He happily ignored the ever-growing media requests. (“You probably won’t be better off after you have appeared on TV than before,” he said.) He tossed out as many invitations, unopened, as he acknowledged. None of this arose from a sense of modesty. Amos knew his own value. He didn’t need to make a point of not caring what people thought of him; he actually just didn’t care all that much. The deal Amos offered the encroaching world was that their interaction was to be on his terms.

  And the world accepted the deal. United States congressmen called him for advice on bills they were drafting. The National Basketball Association called to hear his argument about statistical fallacies in basketball. The United States Secret Service flew him to Washington so that he could advise them on how to predict and deter threats to the political leaders under their protection. The North Atlantic Treaty Organization flew him to the French Alps to teach them about how people made decisions in conditions of uncertainty. Amos seemed able to walk into any problem, however alien to him, and make the people dealing with it feel as if he grasped its essence better than they did. The University of Illinois flew him to a conference about metaphorical thinking, for instance, only to have Amos argue that a metaphor was actually a substitute for thinking. “Because metaphors are vivid and memorable, and because they are not readily subjected to critical analysis, they can have considerable impact on human judgment even when they are inappropriate, useless, or misleading,” said Amos. “They replace genuine uncertainty about the world with semantic ambiguity. A metaphor is a cover-up.”

  Danny couldn’t help but keep noticing the new attention Amos was receiving for the work they had done together. Economists now wanted Amos at their conferences, but then so did linguists and philosophers and sociologists and computer scientists—even though Amos hadn’t the faintest interest in the PC that came with his Stanford office. (“What could I do with computers?” he said, after he’d declined Apple’s offer to donate twenty new Macs to the Stanford Psychology Department.) “You get fed up with not being invited to the same conferences, even when you would not want to go,” Danny confessed to the Harvard psychiatrist Miles Shore. “My life would be better if he weren’t invited to so many.”

  In Israel, Danny had been the person real-world people came to when they had some real-world problem. The people in real-world America came to Amos, even when it wasn’t obvious that Amos had any reason to know what he was talking about. “He had a hell of an impact on what we did,” said Jack Maher, who was in charge of training seven thousand pilots at Delta Air Lines when he sought Amos’s help. In the late 1980s, Delta had suffered a series of embarrassing incidents. “We didn’t kill anyone,” said Maher. “But we’d had people getting lost, people landing at the wrong airports.” The incidents nearly always could be traced back to some bad decision made by a Delta captain. “We needed a decision model and I looked for one, but they didn’t exist,” said Maher. “And Tversky’s name kept popping up.” Maher met with Amos for a few hours and told him his problems. “He started speaking in math,” said Maher. “When he got into linear regression equations I just started to laugh, then he laughed, and stopped doing it.” Amos then explained, in plain English, his work with Danny. “He helped us to understand why pilots sometimes made bad decisions,” said Maher. “He said, ‘You’re not going to change people’s decision making under duress. You aren’t going to stop pilots from making these mental errors. You aren’t going to train the decision-making weaknesses out of the pilots.’”

  What Delta Air Lines should do, Amos suggested, was to change its decision-making environments. The mental mistakes that led pilots of planes bound for Miami to land boneheadedly in Fort Lauderdale were woven into human nature. People had trouble seeing when their minds were misleading them; on the other hand, they could sometimes see when other people’s minds were misleading them. But the cockpit culture of a commercial airliner did not encourage people to point out the mental errors of the man in charge. “Captains at the time would be complete autocratic jerks who insisted on running the show,” said Maher. The way to stop the captain from landing the plane in the wrong airport, Amos insisted, was to train others in the cockpit to question his judgment. “He changed the way we trained pilots,” said Maher. “We changed the culture in the cockpit and the autocratic jerk became no longer acceptable. Those mistakes haven’t happened since.”

  By the 1980s, the ideas that Danny and Amos had hatched together were infiltrating places the two had never imagined them entering. Success created, among other things, a new market for critics. “We started this unknown field,” Amos told Miles Shore in the summer of 1983. “We were shaking trees and challenging the establishment. Now we are the establishment. And people are shaking our tree.” Those people tended to be self-se
rious intellectuals. Upon encountering Danny and Amos’s work, more than a few academics experienced the sensation that a person feels when a total stranger walks up and begins a sentence, “Don’t take this the wrong way, but. . . .” Whatever might follow, you just know that you’re not going to like it. The sound of laughter coming from the other side of Amos and Danny’s closed door hadn’t helped. It caused other intellectuals to wonder about their motives. “The glee is what created the suspicion,” said the philosopher Avishai Margalit. “They looked like people standing in front of a monkey cage, making faces at the monkeys. There was too much joy there. They said, ‘We’re monkeys, too.’ But no one believed them. The feeling was that the joy that they have is to trick people. And it stuck. It was a real problem for them.”

  At a conference back in the early 1970s, Danny was introduced to a prominent philosopher named Max Black and tried to explain to the great man his work with Amos. “I’m not interested in the psychology of stupid people,” said Black, and walked away. Danny and Amos didn’t think of their work as the psychology of stupid people. Their very first experiments, dramatizing the weakness of people’s statistical intuitions, had been conducted on professional statisticians. For every simple problem that fooled undergraduates, they could come up with a more complicated version to fool professors. At least a few professors didn’t like the idea of that. “Give people a visual illusion and they say, ‘It’s only my eyes,’” said Princeton psychologist Eldar Shafir. “Give them a linguistic illusion. They’re fooled, but they say, ‘No big deal.’ Then you give them one of Amos and Danny’s examples and they say, ‘Now you’re insulting me.’”