Identity Crisis

My identity was stolen in 2003 in an entirely low-tech way—from the chart hanging on the end of my hospital bed. The thief rented apartments, ran up utility bills, opened credit card accounts, and rented ski condos using just my Social Security number; he didn’t even have my correct name or my address.

He stole about US $50 000 and was never caught. It took my wife and me over two years of continuous work and about $20 000 in legal fees and forfeited income to get this all off our credit cards.

The biometric solutions Anil K. Jain and Sharathchandra Pankanti propose [“A Touch of Money,” July] might help with the credit card issue, but until the Government prohibits the use of your Social Security number for driver’s licenses, Medicare IDs, medical IDs, and a multitude of other things, identity theft will remain an easily committed and seldom-punished crime.

Robert D. Parker

IEEE Senior Member

Reno, Nev.

Jain and Pankanti would like to see fingerprint sensors everywhere. I would like to see them, and all biometric sensors, banned. The last thing anyone needs is to have a finger cut off or an eye cut out by some person stealing a credit card or a car—as has already happened in Malaysia (see http://news.bbc.co.uk/2/hi/asia-pacific/4396831.stm).

Even if the sensor can distinguish between a live body part and one recently detached from its owner, that won’t stop the thieves from trying. No property or bank balance is worth losing a body part for!

Alan Chattaway

IEEE Member

Surrey, B.C., Canada

Really Robotic

Having read “Robots Can Ape Us, But Will They Ever Get Real?” [Spectral Lines, July], I’ll argue that greater-than-human intelligence either won’t happen or will happen rather a long time from now. Carnegie Mellon roboticist Hans Moravec basically suggests that computers will become intelligent once the hardware is sufficiently fast. Google’s 100 000-node distributed computer should be close to having processing power equivalent to a human’s; it should therefore show signs of self-awareness if Moravec is right.

Maybe the Google cluster is a bit small, but throughout the world we have a network that is on the order of a billion processors (more if we include mobile phone handsets). These are fairly well integrated via things like SETI@home and peer-to-peer file sharing, address book synchronization, Alexa toolbars, and so on. So if computers can become self-aware, the Web should be starting to show signs of self-awareness by now.

Chuck Simmons

Redwood City, Calif.

Concerned Reader

The sidebar “The Big Data Bang” in “Old World, New Grid” [July] states that at CERN the “largest of the detectors, ATLAS, fills an underground cavern over six stories high. It is expected to detect 1 billion collisions per second.”

This detection rate seems improbably high on several grounds, but I will mention just one. Even at the speed of light, particles from one collision will be able to travel only 30 centimeters on average before those from the next collision are generated, creating a number of “waves“ of particles in a large detector simultaneously. Even if all particles travel at close to the speed of light, so that the waves do not overlap, how are the various events going to be isolated?

Bruce Carsten

IEEE Member

Corvallis, Ore.

Authors Fabrizio Gagliardi and François Grey respond: The numbers in the article are correct, but reader Carsten is very astute. Indeed, although there are a billion collisions per second on average, the protons travel in bunches, so the detectors are read out “only” 40 million times per second, which is the bunch-bunch collision frequency. There are then a series of online filters that sort out on the fly whether a given bunch-bunch collision contains an event of significance. Most of the time, this is not the case, so the data are stored at “only” about 100 events per second. Indeed, even at this rate, the large majority of the events stored are background. It is estimated that of the 8 million or so events that will be stored every day for further analysis on the international EGEE grid (http://www.eu-egee.org), on average just one will contain the trace of a Higgs boson—if the Higgs boson exists!

Shedding Light

In his Reflections on “Famous People” [July], Robert W. Lucky stated that “every person would know that Edison invented the electric light.” Though a majority of people might think that Edison invented the electric light, he did not. Electric light was developed in the form of arc lamps roughly 70 years before Edison improved the electric lightbulb, which, just to be clear, he did not invent either.

Sebastian Stüker

IEEE Student Member

Karlsruhe, Germany