The Undoing Project (+ Extra re: new videos)

The Undoing Project


Leonard Zwelling

This is the title of a new book. It was written by Michael Lewis, the best-selling author of Moneyball and The Big Short along with many books that have in common the explication of complexity.

The new book tells the story of Danny Kahneman and Amos Tversky. The two Israelis teamed up to discover how the human mind really works. One of them gained the Nobel Prize in economics for the work. The other, unfortunately, had died at the time of the award so could not share it as the Nobel Prize is not bestowed posthumously.

What is most amazing is that Kahneman and Tversky were not economists. They were psychologists who probed the way humans think, reason, predict, decide, and value their choices. To fully grasp what they uncovered, a series of “heuristics” or rules of thumb about human logic, one must read the book or the original papers the two wrote. Suffice it to say that we humans do not think logically. We do not weigh reality as we think we do and we are not innately Bayesian in our approach to statistics or probability, although we should be. We believe small samples represent bigger realities when they simply may be noise in the system. People predict things by making up stories. We fail to foresee things that in retrospect appear inevitable. Tversky describes it as handwriting on the wall, but in invisible ink. People make decisions to minimize regret. And, my favorite, Tversky said, “we study natural stupidity instead of artificial intelligence.”

Cut me a slice of that!

Obviously, like all of Lewis’ books this is a masterpiece of reducing the complex to the comprehensible. But what does any of this have to do with reality?


We need some help in understanding how the people of the United States and the system they use to elect leaders spit out Donald Trump as its president-elect. I think the innate workings of the human mind as described by Hahneman and Tversky may hold a bit of the answer.

Essentially Americans thought there would be less regret in having Trump in the White House than Mrs. Clinton. Admittedly, the popular vote did go Mrs. Clinton’s way so this analysis may be wanting, but under the rules by which we elect a president, Mr. Trump’s strategy worked and Mrs. Clinton’s did not.

Mr. Lewis came to learn about Kahneman and Tversky after others mentioned how their work really was the basis for the rules that Billy Bean had used in revamping the Oakland Athletics as described in Moneyball. Essentially Bean redefined the rules of the game and hired accordingly. Trump redefined what it means to campaign for President. Kahneman and Tversky redefined the rules of human decision making in the presence of uncertainty. In essence, they originated the concept of evidence-based medicine by showing that so much of what doctors did was based on suppositions that when tested turned out to be wrong.

I think this also explains how poor leadership arises. Systems, be they democratic or autocratic, select people to lead on the basis, not of their proven leadership abilities, but on the assumption that their expertise in one area (e.g., molecular biology) will translate into high performance in another (e.g., leading a major cancer center).

The solution to the problem of faulty human judgment and undependable “gut reactions” is algorithms constructed to generate the desired outcome guiding the human decision-making.

Neither the American political system nor the manner in which academic leaders are chosen adhere to the principle that the system of selection is likely to produce a good result.

As Kahneman and Tversky demonstrated repeatedly, our inferences cannot be trusted for a host of reasons. The neuroscience behind those reasons has yet to be worked out, but the results in the real world are clear.

On a single scrap of paper from 1972, Tversky wrote the following:

People predict by making up stories
People predict very little and explain everything
People live under uncertainty whether they like it or not
People believe they can tell the future if they work hard enough
People accept any explanation as long as it fits the facts
The handwriting was on the wall, it was just the ink that was invisible
People often work hard to obtain information they already have and avoid new knowledge
Man is a deterministic device thrown in a probabilistic Universe
In this match, surprises are expected
Everything that has already happened must have been inevitable

These are important issues in politics and in academic medicine. Predicting performance is predicting outcomes and predicting outcomes is both hard and fraught with error. Yet, predict we must as citizens and as physicians.

We vote by predicting which candidate is most likely to do what we think is a good job. Ditto groups charged with selecting the leaders of academic institutions. This is a very error-prone activity when done by humans as we in America and in Houston now know.

EXTRA: See new videos from book signing and short ObamaCare talk on under Red Kool-Aid, Blue Kool-Aid tab

Leave a Comment

Your email address will not be published. Required fields are marked *