Unlearning Lesson #16
Question #16: Which of the following characteristics has a higher correlation to the success of a Hollywood movie: the involvement of a famous movie star or the location(s) where the movie is shot?
The answer, to the surprise and chagrin of many Hollywood producers, is the latter. The fact that Johnny Depp, Angelina Jolie, Jack Nicholson or any other Hollywood “A” list actor performs in the movie only has a small correlation to its ultimate success. As Ian Ayres recounts in his book, Super Crunchers, a company, Epagogix, now uses a proprietary neural network equation that relies on little more than the information found in a movie’s screenplay to predict—with a higher level of accuracy than Hollywood’s most experienced directors and producers—how financially successful the movie will be. Not surprisingly this finding has not been warmly received by most Hollywood elites because it highlights how little they actually know about the characteristics which truly matter most to a movie’s bottom-line.
This habit highlights another behavior individuals may need to unlearn as they move into the future, and this is the idea of assigning too much weight to one’s own opinions or intuition. Extraordinary new capabilities in data mining and computer processing power, when paired with equally powerful advances in algorithms, neural networks and predictive analysis software, are proving that machines are frequently far better judges than humans at many important tasks.
For example, researchers have found that for patients with uncommon medical conditions Google is now more reliable at assessing those conditions than doctors. Considering that the average doctor has access to only two million pieces of medical information and Google a thousand times more, this isn’t that surprising. What is worth considering is the idea that as computer hardware and software improves exponentially so too will Google’s ability to diagnose a growing number of diseases with an even higher rate of accuracy.
In the rarified field of oenology (the study of wine), Orley Ashenfelter has produced a sophisticated regression equation that can predict more accurately than the wine industry’s top judges which vintages of red wine will be the most valuable in the future. Ashenfelter cares nothing for swishing wine around in his mouth in search of subtle hints of oak, cherry, tobacco or blueberry and instead relies on weather-related data.
For wine collectors who invest in wine futures, Ashenfelter’s information is giving them a distinct competitive advantage over those who can’t or won’t unlearn their reliance on so-called wine “experts.” And only recently have many Las Vegas hotels and casinos unlearned the idea that its most valuable customers are limited to the ranks of wealthy, high-rollers. Thanks to complex algorithms hundreds of middle-aged, middle-income individuals are now being aggressively courted. If you’ve ever wondered why your ne’er-do-well Uncle Ned is being “comped” to stay at a nice hotel in Las Vegas twice a year, it’s because a computer program has figured that it is likely to make the most money off him. (For a reminder of why Ned is so gullible review Unlearning Lesson #9).
Therefore, if you ever happen to receive a free weekend in Las Vegas courtesy of a hotel-casino, my advice is to bet against yourself and stay home. It’ll save you money.
Homework assignment #16: Recalling that it took Galileo years to convince people that light objects fell as quickly as heavy objects even though they could see the results with their own eyes, how long—if ever—do you think it will take before all doctors will be required to confirm their diagnosis of patients with a machine? Defend your answer.
P.S. If you would like to read 38 additional “unlearning lessons,” consider picking up a copy of my new book, Higher Unlearning: 39 Post-Requisite Lessons for Achieving a Successful Future. The eBook is now only $2.99!