"Any sufficiently advanced technology is indistinguishable from magic," the science-fiction writer Arthur C Clarke once observed. Many technologies do seem supernatural the first time you see them. Microwave ovens heat up food using invisible rays; music can be encapsulated on shining disks; computers can simulate vivid, three-dimensional worlds. The technology industry is driven by our desire to be amazed by gadgets. So it is hardly surprising that wild claims are made about new technologies. The idea that a gadget can make your troubles vanish is seductive.
This column (a new regular in Prospect) sets out to examine such claims, not merely to puncture hype, but also to sing the praises of technologies that are indeed changing the world, often hidden from view. It will look at why some technologies succeed while others, though promising, fail to get off the ground. Can genetically-modified rice feed the developing world? Can desalination technology prevent a water crisis? Can new kinds of scanning equipment rid the world of landmines?
Whether a technology succeeds or fails usually has as much to do with external factors-political, social or economic-as with the technology itself. Consider a technology that has been the subject of much interest since 11th September: biometrics, or body scanning. There are numerous biometric technologies, including fingertip scanning, hand-geometry analysis, iris and retina scanning and facial recognition. In each case, a measurement of some property of an individual's body is compared to stored measurements known to be from particular people, either to verify that a person is who he claims to be, or to see if he is on a list of wanted individuals. Such systems are used to secure bank vaults, to ensure the right people are let in and out of jails, and to grant access to computer centres.
But the technology has never found a mass-market application. Back in the 1980s, it was touted as the best way to secure office computer networks from hackers, but passwords proved cheaper to implement than providing every PC with a fingerprint scanner. In the 1990s, biometrics firms tried to capitalise on concern about the dangers of sending credit-card information over the internet. But buying things online, with your financial details protected by just a password, turns out to be safer than using your credit card in a dodgy restaurant. Last year, the biometrics vendors latched on to internet shopping on mobile phones: only biometric scanning, they suggested, could prove that the person using the phone was its owner. But enthusiasm for mobile commerce fizzled, so it was back to the drawing board. Have the biometrics firms finally struck gold, by rebranding their wares as anti-terrorist tools?
Those in the industry certainly seem to think so. Tom Colatosti of Viisage, a leading biometrics firm, has claimed that facial-recognition systems (of the kind that his own company sells) could spot terrorists before they board aeroplanes. The events of 11th September have prompted a dramatic tightening of airport security procedures, and interest in the products made by biometrics firms has surged. Another biometrics company, Visionics, had to install new phone lines to cope with the demand, and raised $20m in additional funding within days of the attacks. When the New York stock markets reopened the share prices of both Viisage and Visionics more than doubled.
Time for a pinch of salt. The problem is that all the biometric systems in the world would not necessarily prevent a repeat of the 11th September hijackings. Although some of the hijackers travelled under false identities, others appear to have used their own names; so biometric signatures incorporated into passports, to make them harder to forge, would not have stopped them. What about the idea proposed by Visionics, that passengers should submit to a face scan? This is only any use if the terrorists are already known to the authorities and their pictures are in the suspect database. In the case of these hijackers, only two were on the FBI's watch list. Biometrics cannot make up for the failings of the intelligence services.
Another problem is that no biometric system is 100 per cent reliable. Even a system that is 99.99 per cent accurate will produce a vast number of "false positive" matches when faced with tens of millions of travellers. It may correctly identify a terrorist or two, but it will incorrectly finger thousands of innocent people. Meanwhile, libertarians worry that biometric surveillance undermines civil liberties, and warn of "function creep": a system installed to look for terrorists might end up being used to catch people with overdue parking tickets.
There is much scope to improve airport security. But the most effective measures involve procedural changes-such as restricting access to secure areas and improving the status of baggage screeners. The adoption of biometrics would simply move the weakest link in the security chain to somewhere else; ultimately, security must be enforced by people, not machines. That is not to say that biometrics is entirely without merit. But it is no silver bullet. No technology ever is. n
