Tales of the unexpected

Originally published in Global Agenda, January 2003

Today’s conventional wisdom is as certain to be wrong as the conventional wisdoms of the past. Think of the end of history, how dotcoms were going to destroy incumbents, Dow 36,000 and global financial architecture. The Carnegie Endowment’s Foreign Policy magazine recently devoted an entire issue to “bad ideas”, such as the dependency theory of under-development, the superiority of Japanese management, the Laffer curve or — on a far grander scale — Marxism.

And we’re certainly bad at understanding the long-term significance of events. When asked about the significance of the French revolution, Chinese revolutionary leader Zhou Enlai replied, “It’s too early to say.” But you don’t even need that kind of perspective to understand our failings.

To take just one example: in July 1969 the first man walked on the moon; six weeks later the first use was made of an embryonic Internet. One event held the world in awe, but has so far resulted in little tangible benefit for the world. Perhaps a dozen scientists and engineers were aware of the second event, but it has provoked profound change. Even if more people had been aware of the Internet transmission, it is certain they would have had no idea about its consequences.

We’re even bad at forecasting some very near-term events. Think back 15 or even ten years. Few expected a civil war in Europe with hundreds of thousands killed, half a billion users of the Internet, a long stagnation in Japan and a long boom in the United States, or the complete mapping of the human genome. Even more recently, a Davos session on Afghanistan in 2001 attracted one of the smallest audiences of the Annual Meeting.

What makes us so bad at understanding or forecasting the course of events?

First, we are understandably preoccupied with the events and facts that loom large at the moment. There is a tyranny of the immediate. Try to remember the last time, in response to a problem, someone told you, “I’ll sleep on it.” Our organisations demand rapid judgements, quick responses. In business, despite the many counter examples, it is still generally believed, as one-time corporate hero Percy Barnevik said, “It is better to be fast and wrong, than slow and right.”

This tendency is exacerbated today by 24-hour media and the constant response of the markets. So we are preoccupied with Afghanistan, then Iraq, then North Korea. Or Enron is succeeded by Global Crossing, by Tyco, by Imclone. Who has the time or energy to look at what’s roiling beneath the surface?

We are further hampered by tunnel vision and over-enthusiastic extrapolation, what John Seely Brown and Paul Duguid in The Social Life of Information term the “1, 2, 3 — a million” tendency.

Tunnel vision makes us see an uninterrupted, clear road ahead, where events proceed apace (as memorably pictured on the cover of Bill Gates’s The Road Ahead). But events rarely advance the way we foresee. And as roads ahead become bumpy and twisty, extrapolative leaps become highly inaccurate. Responsible economists are fond of graphs showing how forecasts scatter and diverge over time, as extrapolation of existing data becomes less reliable. But too many of us prefer the simplicity and certainty of the single-line charging up the graph.

So with these caveats in mind, and in an attempt to zig on some issues where the majority is zagging, here are three of the less expected events that we should be focusing on for 2003.

The privatisation of surveillance

Terrorist outrages have made the balance between security and civil liberties a key issue in many countries. The US government’s Total Information Awareness programme, under the direction of John Poindexter, is perhaps the most ambitious example. TIA will create a virtual, centralised grand database that will, as reported in The New York Times, “provide intelligence analysts and law enforcement officials with instant access to information from Internet mail and calling records to credit card and banking transactions and travel documents, without a search warrant”. Unsurprisingly, TIA has attracted fire from civil libertarians on both left and right.

But there is relatively little attention being paid to a parallel development, which is the spread of private means of surveillance. “I think the Drudge Report will meet the x10 cam,” says Jonathan Zittrain, director of the Berkman Center for Internet and Society at Harvard Law School. “Thanks to cheap wireless connections, the public will start being able to survey almost anything of interest (and much of no interest) in Western countries’ public areas. Videos like the Rodney King beating will become a commonplace, streaming out as they happen.”

Zittrain points to a website at the University of Pittsburgh where users can control a series of cameras on the campus’s public spaces. More piquantly, some commentators, irate at the TIA plans, have posted publicly available aerial photos of Poindexter’s house. The curious can see the house only a stone’s throw from a green at Lakewood Country Club.

Zittrain’s vision of real-time streaming videos from all sources will only need the next small step in mobile telephone technology, as the picture-enabled phone evolves into the video-enabled phone.

The privatisation of surveillance has already been taken into space. High-resolution satellite images were the exclusive province of a handful of intelligence agencies only a couple of years ago. But now companies like Orbimage and Spin-2 provide all-comers with satellite images. Through a site like Terraserver.com, you can browse the world. The skills that enable the CIA to identify an al-Qaeda training camp may not be readily available, but the raw information is not far out of reach.

Allowing algorithms to kill

On November 3 last year a CIA agent authorised an unmanned drone flying over the Yemeni desert to fire a missile. The car carrying Qaed Salim Sinan al-Harethi, suspected of the 2000 bombing of the warship USS Cole, and five other alleged al-Qaeda terrorists was destroyed.

That operation relied on the sophistication of video technology, transmitting images to an operator hundreds of kilometres away. A human made the decision to fire, even though he was vastly remote from the action. But we’re now on the verge of humans being taken out of the decision.

The next generation of unmanned aircraft will have significantly enhanced capabilities. The interest of the Pentagon in these systems is understandable: they are expected to cost less and, by definition, have no aircrew at risk.

Boeing’s Unmanned Combat Air Vehicle (UCAV), for example, is expected to cost 65% less to produce than future manned fighter aircraft, and 75% less to operate and maintain. What will UCAVs do? “In a typical mission scenario,” Boeing’s literature explains, “multiple UCAVs will be equipped with preprogrammed objectives and preliminary targeting information from ground-based mission planners. Operations can then be carried out autonomously, but can also be managed interactively or revised en route by UCAV controllers should new objectives or targeting information dictate.”

How far are we willing to allow intelligent systems to make lethal decisions? How much autonomy should be allowed?

For example, with improving visual recognition systems it should be technically possible for a UCAV or equivalent to “see” enemy troops and distinguish them from friendly troops either through uniform markings or possibly because future troops will have some unique electronic marker embedded in their uniform. (The issue is already active in military circles because US allies are far behind in development and deployment of combat identification technology, posing potential problems in coalition actions.)

Should unmanned machines be able to choose to shoot enemy troops? How would recognition systems judge surrender? How happy are we to leave lethal decisions to a series of algorithms? Will we reach a point where our faith in the consistency of algorithms supplants our belief in trained human judgement?

Before the technology gets to the battlefield, it may be necessary for states to consider such questions. There is precedent for successful pre-emptive action. The 1967 Outer Space Treaty requires signatories “not to place in orbit around the Earth, install on the moon or any other celestial body, or otherwise station in outer space, nuclear or any other weapons of mass destruction”. Perhaps there is a need for a Robotic Weapons Treaty.


The current geopolitical instability makes it particularly difficult to know where to look. The last year has seen heightened tensions between the two south Asian nuclear powers over Kashmir, admissions from North Korea about the extent of their nuclear weapons programme, continued political and economic instability in South America, the spread of terrorism to east Africa, without even referring to the Middle East.

But some of the most disturbing news of last year came from none of these places. In November more than 200 people in Nigeria died because of rioting associated with the Miss World beauty pageant. Most reporting of the riots in Kaduna suggested they were provoked by a newspaper article which offended Muslims by suggested the prophet Mohammed might have married a beauty queen. But the riots didn’t start until four days after the publication, and reporters who pursued the story (after the mass of media had left) have written they could find few people in Kaduna who knew about the article or had even heard of the Miss World pageant. The tensions were there and needed little spark.

For those of us who live in the wealthy North, news of Nigeria is relatively hard to come by. Although it has a vastly higher population than South Africa or Egypt and is a major oil exporter, it gets far less coverage than either of the other two major African countries (Nigeria is the world’s ninth most populous country, but only the 51st largest economy).

Only 12 of Nigeria’s 31 states follow sharia law. But president Olusegun Obasanjo needs votes from these states — which tend to vote en bloc — to win re-election in April this year. So many of the hopes that Obasanjo’s democratic election four years ago would help transform Nigeria have been frustrated as he placates the Islamic north. Tensions in the north between the Muslim majority and the Christian minority have been worsening for at least two years, with the worst outbreak occurring in February 2000, when over 2,000 people were killed.

Nigeria’s economy relies overwhelmingly on oil from the south (reminiscent of economist Jeffrey Sachs’ quip that Venezuela was an oil company with 24 million employees). For the moment, tensions between the south and north are somewhat allayed by the presence of the southern Obasanjo in the presidency.

Some historically aware observers have compared the way in which northern politicians are using sharia to solidify their base, much as politicians in the US south states used race during the Jim Crow era (in both cases, there is belief as well as calculation involved). It’s not inconceivable that Nigeria’s federal structure will come under strain just as the US one did 150 years ago. For states’ rights, substitute sharia law.

Groups like Amnesty International are forecasting that violence will increase in the months leading up to the April elections. Undoubtedly western media will fly in to Nigeria for the elections, but the inherently fragile situation in Africa’s most populous country needs attention before then.

Leave a Reply

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