by David Orrell
Praise for The Future of Everything
“With wit, humor, and clarity The Future of Everything gives a good overview of the history of science as a predictive tool. . . . Orrell’s work has the potential to change the way we plan for the future, both personally and as a society.” —Winnipeg Free Press
“Even big-picture predictions—global warming, impending pandemics, the global impact of consumerism, emerging technologies such as genetic engineering and nanotechnology—seem less overwhelming after this enlightening history lesson. . . . Some of Orrell’s most eloquent and provocative passages are about those unpredictable forces known as free will and personal choice, and he rallies us all to effect positive change.” —Canadian Geographic
“If you loathe uncertainty, can’t cope with complexity, or prefer easy (yet wrong) solutions to complex problems, you may not appreciate this book. On the other hand, you will likely find it reassuring, even empowering perhaps, to know that the ‘experts’ can be wrong, and that common sense is still useful.” —Edmonton Journal
“The special and enchanted land of forecasting is inhabited by tribes of specialists who speak a language that is unintelligible to most of us. David Orrell’s The Future of Everything is a wonderful guide and companion to that land; its history, its valuable resources and its fault lines. And it still manages to be a good read.” —Everett Herald
“Orrell’s writing is top-notch, but he’s at his finest when writing about the old days, somehow finding exactly the right mix of anecdote, broad brush, and humor. There’s the goat’s-beard hygrometer (curly plant = humid weather) and the invention of the word ‘forecast’ (to distance weather prediction from astrological prediction, though the same folks often did both).” —The Globe and Mail
“Mathematician David Orrell explains why the mathematical models scientists use to predict the weather, the climate, and the economy are not getting any better, just more refined in their uncertainty . . . Dr. Orrell is no climate-change denier. He calls himself green. But he understands the unjustified faith that arises from the psychological need to make predictions.” —The National Post
THE FUTURE OF
EVERYTHING
THE FUTURE OF
EVERYTHING
THE SCIENCE OF PREDICTION
FROM WEALTH AND WEATHER
TO CHAOS AND COMPLEXITY
DAVID ORRELL, PHD
The Future of Everything
The Science of Prediction
From Wealth and Weather to Chaos and Complexity
Copyright © 2007 by David Orrell
Hardcover edition published in January 2007 in the United States by Thunder’s Mouth Press, by arrangement with HarperCollins Publishers Ltd., Toronto, Canada
Paperback edition published in December 2007 in the United States by Basic Books, A Member of the Perseus Books Group, 387 Park Avenue South, New York, NY 10016-8810
All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system now known or to be invented, without permission in writing from the publisher, except by a reviewer who wishes to quote brief passages in connection with a review written for inclusion in a magazine, newspaper, or broadcast.
Library of Congress Cataloguing-in-Publication Data is available.
Hardcover: ISBN-10: 1-56025-975-2; ISBN-13: 978-1-56025-975-6
Paperback: ISBN-13: 978-1-56858-369-3
eBook ISBN : 9780786732319
9 8 7 6 5 4 3 2 1
Printed in the United States of America
This book is dedicated to the memory of my father, John Orrell, to Beatriz and to Isabel
INTRODUCTION
THE SCIENCE AND SOCIOLOGY OF FORECASTING
The term “natural disaster” has become an increasingly anachronistic misnomer. In reality, human behavior transforms natural hazards into what should really be called unnatural disasters.
—Kofi Annan, Secretary-General of the United Nations
Prediction is very difficult, especially if it’s about the future.
—Niels Bohr, Nobel laureate in physics
ANATOMY OF A STORM
By December 15, 1999, the run-up to the new millennium had begun in earnest. People around the world were getting the champagne in, ready to uncork it the moment 1/1/2000 did its giant, slow march into their time zone. The NASDAQ stock-market index was also acquiring a champagne-like froth, as the Internet rewrote the rules of the world economy. The only cloud on the horizon—a potential storm—was the Millennium Bug, a software error caused when programmers rounded off the computer’s internal date from four digits to two. Many predicted it would bring chaos, or even collapse, to the world economy.
In the slums, or ranchos, perched high on El Avila mountain north of Caracas, Venezuela, the Millennium Bug was not a major worry. Most of the residents had electricity, acquired by tapping illegally into the network, but Internet access was limited, to say the least. They were more excited about that day’s referendum, which their hero, the new president, Hugo Chavez, had called to get the constitution approved. Voter turnout was high, and 78 percent in favour, despite the heavy rains, which had been falling for days and seemed to be getting worse.
Normally, the rainy season lasted only until October, but this year was an exception. The weather was out of synch. Some in the government suggested that the more vulnerable areas of Caracas be evacuated, in case the steep slopes became unstable. Perhaps not wishing to disrupt the referendum, however, the government took no action.
Early the next morning, December 16, the northern side of El Avila, which faces the coastal resorts near the airport, simply gave way. Witnesses said that huge waves of water, six metres high, cascaded down the mountain, carrying away everything in their wake—trees, cars, houses, people. Giant boulders hurtled down narrow gullies, not stopping until they fell into the ocean or smashed into the luxury apartment blocks that lined the coast. One survivor, a young woman, described waking up in the middle of the night to the sound of the water and the rocks crashing down the mountain and people yelling, “The river is coming!”1 The water filled half her house before she could get out the door. Many others weren’t so lucky.
At first, the scale of the disaster was not comprehended. Estimates were for 100 dead, then 500. By December 22, this had grown to 30,000, with perhaps ten times that many left homeless. Many of the bodies were swept out to sea or buried under the mud, and were impossible to recover. Survivors gathered in halls and stadiums, desperately searching for lost relatives. Rescue workers were completely overwhelmed. One told a journalist from the Independent how he had rescued a three-year-old girl: “Every time she saw water, she screamed.”2 Many feared that the lack of drinking water and sanitation would lead to disease outbreaks, magnifying the human impact, but fortunately this did not come to pass.
Almost immediately, the storm was politicized. The president’s opponents castigated him for proceeding with the referendum. On Christmas Day, Chavez distributed gifts to hundreds of orphaned children at the Poliedro sports arena, in Caracas. When a reporter suggested the ex-paratrooper was culpable for the disaster, he replied: “They should shoot me if I have any personal responsibility in this.”3
Could this storm, with all of its social, economic, and medical repercussions, have been predicted? Was it a random, unforeseeable event, or was someone responsible? How about the weather forecasters? The conditions for the storm began to develop in early December, when a cold front encountered a southwesterly flow of moist air, resulting in precipi
tation over the northern coast. There was one week of moderate rain, followed by two days of extremely heavy rain on December 15 and 16. The daily totals recorded at the nearby Maiquetia International Airport for these days were so excessive that, in theory, they wouldn’t be repeated for 1,000 years—truly a millennial storm.4 Almost by definition, such events do not get predicted; no forecaster likes to predict something he has never seen happen.
High above the storm, watching it develop, was a GOES 8 satellite belonging to the National Oceanic and Atmospheric Administration. Its mechanical infrared eye didn’t pick up the horror on the ground; it was focused on the cloud-tops, taking temperature readings that could be used to estimate rainfall. However, the relatively coarse resolution—the smallest features it could detect were roughly four kilometres by four kilometres—meant that it could only confirm the location of the maximum rainfall.5 The storm seemed to stand still, as if it was intent on bringing the mountain down.
Even with the amount of rain, though, no one could have foreseen the scale of the mudslides. And if they had, no one would have paid attention. As a forecaster at the National Hurricane Center in Miami admitted, for every mudslide they got right, they were “going to end up screaming wolf maybe 10 times.”6 The instability of the soil was the result of a number of factors, not all of them natural. The poor rancho areas were completely unregulated by the government; in fact, because of high levels of crime, many were nogo zones for the police.7 Much of the forested land around the hills had been clear-cut by residents for firewood and construction material or cut down by private companies, weakening the soil. Benches had been cut into the steep hillsides to support the houses of cinder block and corrugated iron, further destabilizing the ground. Many of the homes were situated in dry riverbeds, so were directly in the path of the mudslide as it coursed down the gullies.
In fact, the disaster was caused by a range of complex, intertwined forces. Historical records from Spanish archives show that major floods and landslides are hardly new to the area.8 Casualties were so high this time because of the sheer number of people— around 3 million—who now make their homes on the highly inhospitable mountains clustered around Caracas. This is a result not of extreme weather but of extreme disparities in wealth and migration to overpopulated urban areas. The storm, and its unusual timing, might also have been influenced by global warming, which in turn depends on the amount of carbon dioxide emitted from the world economy. Warmer oceans mean more water gets evaporated, so storms are expected to become more powerful.9 The stakes for forecasters may grow higher still.
When I visited the flood area a year later, it was still covered in a layer of light brown soil. Since it was impossible to recover bodies, some of which were buried six metres down or more, the Catholic priests consecrated the entire place as a burial site. Already, though, the ranchos are encroaching on the territory they lost. Was it really a once-in-a-millennium event, or will the same thing happen again in the next hundred years? Or the next ten?
MAPPING THE FUTURE
The Future of Everything is about scientific prediction in the areas of weather, health, and wealth—how we foresee storms or fair weather, sickness or health, booms or crashes. It might seem that forecasts of the atmosphere have little to do with prediction of diseases or the economy, but in fact these three areas are closely linked. For one thing, they often affect each other, so prediction is an intrinsically holistic business. As shown above, a storm’s impact depends on the conditions on the ground, and can have huge economic consequences. When Hurricane Katrina swung into the Gulf coast in late August 2005, flooding much of New Orleans and knocking out oil refineries, its financial impact was greater than that of the 9/11 terrorist attacks. In 2003, for Toronto, Hong Kong, and other cities, the storm was called SARS. Global warming too is a multistranded problem with complex repercussions. Like a potentially larger version of the Caracas storm, it exists at the centre of a vortex of social and environmental causes and effects. Atmospheric carbon dioxide is influenced by economic output and population levels; the resulting climate shifts and environmental stress may affect the spread of disease; large-scale epidemics have in the past severely disrupted economic activity; and so on.
The three types of prediction also use similar methods and share a common past. A traditional technique is astrology, which links the biological event of a baby’s birth or the atmospheric and economic event of good harvest weather to the motion of the planets. Many people begin their day by reading their horoscope in the newspaper. Because I’m a Gemini (and therefore conflicted), I read my horoscope occasionally but never believe it. However, I will gladly check the five-day, long-range weather forecast, which probably has a lower accuracy rate.
For much of history, the same experts supplied both horoscopes and weather forecasts; humans and the atmosphere danced to the same tune. The seventeenth-century astronomer Johannes Kepler paid his way through university, and indeed much of the rest of his life, this way. Even now, when we talk about the weather, we often describe it in almost human terms. It is a cheerful day or a gloomy one; a storm is violent; a hurricane has a mind—and always a name—of its own. The weather is a character in all our lives, and sometimes it is a criminal.
The first newspaper weather map appeared on April 1, 1875, in The Times of London. It was prepared by the British scientist Sir Francis Galton, who also discovered the anti-cyclone.10 Three years later, Galton gave a lecture at London’s Anthropological Institute on a rather different subject. He presented composite photographs of the faces of prison convicts, dividing them into three groups by the type of offence. His aim was to search for common characteristics within each group. Influenced and inspired by the ideas of his cousin, Charles Darwin, Galton believed that traits such as “eminence” and criminality were inherited and linked to physical appearance. Just as he had scanned weather maps for patterns that would foretell a coming storm, he now looked for facial characteristics that warned of criminality. To help bring out such features, he made the composite pictures, which overlaid the faces of up to eight individuals.11
The results were not very useful for prediction. What they showed was more a common humanity than any particular demonic trait. (As discussed in Chapter 5, this didn’t prevent Galton from inventing the field of eugenics to “improve” the human race.) Since Galton’s time, a huge scientific effort has been devoted to looking into our future weather, health, and wealth, now using mathematical models. These emulate the flow of air and water in the atmosphere, or substances in our body, or money in the economy, using large sets of equations. Although the calculations are performed on high-speed computers, the techniques are essentially the same as those first developed by physicists such as Isaac Newton to study the dynamics of celestial objects. Like astrology, our predictive models of the future have their roots in the stars.
Weather prediction has evolved in the past half century into the multi-billion-dollar business of providing up-to-the-minute forecasts to the media and to weather-affected industries, such as agriculture, transport, and insurance. In biology, the Human Genome Project catalogued all human genes in a kind of giant library of our species; one of the stated aims was the prediction and control of genetic traits and diseases. Perhaps the greatest preoccupation of predictors has been the infinitely intriguing motions of the trilliondollar financial markets. Companies, governments, and universities around the world, as well as giant institutions like the International Monetary Fund and the World Bank, hire thousands of economists in an effort to foresee economic events.
It turns out, though, that predictions of weather, health, and wealth have another thing in common. While scientists have had great success in squinting through microscopes at the smallest forms of life, or smashing atoms together in giant particle accelerators to analyze the structure of matter, or using telescopes to look forwards in space and backwards in time at the formation of distant galaxies, their visions into the future have been, like Galton’s composite photogra
phs, blurred and murky. As a result, projections tend to go astray.
In weather forecasting, for example, accuracy has improved in a slow, iterative manner, but if you’re not a natural risk-taker, you should put little faith in a five-day forecast. We can put a man on the moon, but timing a shuttle landing around the weather is still tricky. In medicine, biologists have realized that the connection between genes and traits is not a straight line but a highly twisted and circuitous one. It is frequently announced in the press that the gene that causes some condition has been discovered, only for the news to fade from public attention as the complexities emerge. And in economics, the dominant “efficient market” theory, which has been weirdly embraced by many highly paid predictors, says that, in principle, the economy cannot be predicted. The best one can do is predict and control financial risk. Even this aim seems out of reach after events like the Black Monday crash in 1987 or the bursting of the Internet bubble.
The reason scientists—who are usually not closet Nostradamuses—have been drawn to making predictions in these areas (rather than, say, fashion or popular music) is because the underlying systems seem quantifiable and computationally tractable. Weather is just fluid flow, the human body is biochemistry, the economy is money. So what is going wrong? What do these systems have that escapes the models? Is our difficulty in forecasting the future health of the planet related to our difficulty in predicting the health of our own body? Or the health of the economy? Will we always be blind to the future, reacting impulsively to the next crisis or piece of good fortune when it comes along? Finding the answers to these questions is the target of The Future of Everything.