Waiting for the Revolutions

Don't blame the experts who didn't see the Arab Spring coming.

Blake Hounshell is right ("Dark Crystal," July/August 2011) that "experts," including myself, failed to predict the Arab revolutions. However, there are a few nuances that may make us appear less inept.

First, any revolution, as Charles Kurzman correctly notes, is a low-to-moderate probability event in any given year. We can do a very good job distinguishing more unstable states from very stable ones, but we can't always say which state or states will have a crisis next year. Similarly, a doctor may be able to say that patients who are overweight and have high blood pressure are at a much higher risk of having a heart attack, but can't know which patient is going to have a cardiac episode in the coming year.

More importantly, our forecasting models are capable of being adjusted in response to events. Once the Tunisian revolution occurred (unexpectedly!), we were able to look more closely at conditions in neighboring states and change our forecasts to anticipate a breakdown of order. We were much better able to anticipate the rebellion in Libya, and my May/June Foreign Affairs article, which was originally drafted and circulated at the beginning of March, correctly anticipated the Syrian uprising that began on March 15. It also correctly forecast that the monarchies in Morocco, Jordan, Oman, and Saudi Arabia would remain far more stable than the personalist dictatorships in Tunisia, Egypt, Libya, Yemen, and Syria. So while we may have missed the first onset of events, our analysis was still helpful in other respects.

It would be ideal if everything in the world of politics were nicely predictable. But we are still far from being able to say exactly when and where one will strike.

Professor of Public Policy
George Mason University
Arlington, Va.

Blake Hounshell replies:

Let me first humbly submit that I didn't predict the Arab revolutions either; nor did a number of other quantitative models of instability, including some very expensive systems used by U.S. government agencies. The one prominent individual I've found who claimed to have called Egyptian President Hosni Mubarak's fall ahead of time was Bruce Bueno de Mesquita, a political scientist whose forecasting models are used by … U.S. government agencies. I did not mean to single out Goldstone or his work, for which I have the highest regard.

Indeed, I have no doubt that quantifying political upheaval is a difficult but worthy endeavor, and I wish Goldstone well as he refines his models. In the right hands, they clearly have much to tell us -- and perhaps we'll even be better prepared for the next round of falling dictators, like, say, Belarus's Aleksandr Lukashenko and North Korea's Kim Jong Il.

In the meantime, as Mark Abdollahian, whose Sentia Group does forecasting work for various government agencies, told Wired magazine in February: "Think of this like Las Vegas. In blackjack, if you can do 4 percent better than the average, you're making real money."

Peter Macdiarmid/Getty Images


Love and Robots

Artificial intelligence expert David Levy says relationships with robots might be even more complicated than Ayesha and Parag Khanna assume.

In their lucid and convincing piece, Ayesha and Parag Khanna ("Technology Will Take on a Life of Its Own," September/October 2011) use the example of a young man in Japan who recently married a video-game character, pointing to a future in which human-robot marriages will be commonplace. Yes, we humans will be forming emotional attachments with machines, falling in love with them, having sex with them, and marrying them, thereby contributing significantly to a worldwide reduction in loneliness and the unhappiness that the lack of a loved one so often brings. But there are dangers as well.

The article also touches on what will be one of humankind's greatest challenges in the future -- cyberhacking. The Khannas ask whether it is only human cyberhackers whom we have to fear, "or perhaps also artificially intelligent software programs." I would answer both, and their threat to just about every facet of our lives will be awesome.

In the case of robots built for love and sex, one aspect of what we have to fear from cyberhackers is encountering robots that play with our emotions because their software has been hacked to make them do so. The warped thinking that encourages hackers to wreak havoc with computer systems just for the "fun" of it is likely to instigate, for example, a type of virus that manipulates human emotions sufficiently to make someone fall in love, and then dash their bliss by dumping them. This and other emotional "games" will have the potential to cause misery to the point of making some who love robots suffer extreme psychological trauma.

We need Alvin Toffler-like thinking to prepare us for these threats and point the way for dealing with them.

CEO, Intelligent Toys Ltd.
Author, Love and Sex With Robots
London, England

Ayesha and Parag Khanna reply:

We welcome David Levy's insight, based on his own intellectually stimulating research. He rightly points to the unintended consequences of our growing emotional attachment to machines. While some maladies such as loneliness are addressed, other vulnerabilities are created. As pervasive networks expand -- whether among humans or among machines and humans -- there is a window or lag time in which users can be manipulated. Hybrid Reality Institute fellow Marc Goodman has conjured very specific scenarios around the possibility of financial algorithms' gaining the autonomous capacity to divert capital toward shell companies, effectively stealing from the markets.

Stanford University's Jeremy Bailenson, an expert on virtual-human interactions, predicts that eventually anti-malware software will be developed that alerts us to the kinds of intrusions Levy rightly fears will transpire. But with the stakes and potential profits from online emotional extortion rising (just witness the many fraudulent schemes that have pervaded Second Life), cyberhackers will no doubt persist in their efforts to penetrate not just individual consumers, but en masse. In a related vein, recent reports have documented the rise of "virtual slavery" in which Chinese prison inmates are forced to play online games for up to 12 hours per day (in addition to hard manual labor) in order to "gold-farm" -- build stockpiles of virtual currencies that prison guards then spend. We know all too well how regulations and education frequently lag one or more steps behind such crafty and malicious purveyors of cyberexploitation. It is therefore a task of foresight to anticipate these possibilities.