Like any other conference, it allows people with similar interests to meet and make contact, whether for academic collaboration, business partnership or simply to make friends. But this process of “networking” is usually rather mysterious. Place a large number of people in a room, add some food and alcohol, and somehow it happens.

But at Pop!Tech this year, things were different. As delegates arrived at the meeting, they were handed an intelligent tag the size and weight of a PDA to wear around their necks. Called an nTag, each delegate’s device was pre-programmed with the conference schedule, which could be displayed on a small screen on the front of the tag, as well as with personal information supplied earlier to the organisers.

This included the wearer’s contact details, employment history, their professional interests and personal hobbies the kind of information that people often compare to decide whether they have anything in common.

The purpose of nTags is to ask all those ice-breaker questions automatically. The tags communicate with each other via an infrared link to find out whether their owners have much in common. When an nTag finds a good match, it does what any good party host would do and alerts its owner to the other person.

Good host

“I’m really passionate about good conversations,” says Rick Borovoy, co-founder of nTAG Interactive, the New York-based company that hires the devices to conference organisers. The tags are the product of his doctoral work at MIT’s Media Lab in the 1990s, which he launched as a business earlier in 2003.

“We’re really trying to decrease the awkwardness of a meeting,” says Borovoy, “and tell you something that a good host would tell you about the two of you.”

Data and alerts appear on the small monochrome LCD screen of the nTAG. The relatively simple machines run on four AAA batteries and have 128 kilobytes of RAM, and 64 kilobytes of flash memory about enough to store 60 pages of text as an electronic file.

As well as communicating with each other via infrared links, nTags also use an RFID (radio-frequency identity) chip to talk to a central server. This allows delegates to download other people’s details to an email address of their choice. At the same time, the central server updates information held on the tags, such as changes to the conference schedule.

The service costs between $40 and $100 per tag per day, depending on the length and complexity of an event, but that includes peripheral equipment such as the central server and any customised services designed to get people mixing. For example, Borovoy has developed a cocktail party game that gets people to hunt for each other, and it can be adapted in various ways to suit the occasion.

Hacked to sleep

Most of the 488 Pop!Tech delegates were impressed with their nTags, exchanging 3965 visual business cards during the event. But not all were enamoured. Some felt the devices were too heavy, or that the constant interruptions became intrusive, especially after the first day.

Whitfield Diffie, an engineer at Sun Microsystems Laboratories in Palo Alto and the man behind the concept of public key cryptography, felt that the devices were an invasion of privacy. He created a stir by hacking into his nTAG to put it into sleep mode. And to the delight of some delegates and the frustration of others, he set up his device to do the same to any other nTAG it talked to.

Borovoy admits there are some problems and points out that the company is still learning how best to put the tags to use. But he hopes to make nTags more wearable in the form of ties or T-shirts, and to add features such as instant messaging and real-time data updates. “We see nothing but more potential,” he says.

For several years seismologists have intensely debated whether or not you can forecast earthquakes by the electromagnetic signals emitted by rocks under stress.

Proponents of the idea – known as the VAN method after the initials of the Greek researchers who developed it, Panayiotis Varotsos, Kessar Alexopoulos and Konstantine Nomicos, all of the University of Athens – maintain that electrical and magnetic activity in the ground can predict the location, time and magnitude of some earthquakes.

Other researchers have had difficulty repeating the Greek team’s results. But now Seiya Uyeda’s team from the Earthquake Prediction Research Center at Tokai University in Japan reports measuring anomalous changes in the Earth’s electrical and magnetic fields in Japan’s Izu islands, from the end of March 2000. And a couple of months later, a series of earthquakes started on 26 June.

The researchers used telephone wires as antennas, to measure extremely low frequency electromagnetic waves every 10 seconds.

Rock faults

They say their geoelectrical field readings showed “clear, unusual changes”. The strength of the signals increased over time, reaching a peak just before the first large earthquake of magnitude 6.4 on the Richter scale on 1 July 2000. After the seismic activity died down, the fields returned to normal.

The group also saw changes in the Earth’s magnetic field strength over the same period. After allowing for other possible causes of magnetic noise, such as rainfall or man-made sources, the researchers saw a tiny but unexplained distortion. The variations were about a million times smaller than the Earth’s natural magnetic field.

The signals were picked up at only two antennas, which the researchers admit is puzzling. They think that the stress signals may be propagated along highly conductive channels, such as faults in the rock that contain pools of water, making them detectable only in certain areas.

“Grasping at straws”

Seismologists are divided over the significance of the results. Varotsos says that Uyeda’s experimental results are impressive, and confirm the signals he has seen in Greece over the past 20 years.

Max Wyss of the University of Alaska Fairbanks, on the other hand, dismisses the Japanese measurements. The group’s assertions that these signals predict earthquakes are “grasping at straws”, Wyss says. “They simply want to believe that, no matter what the facts are.”

Phil Reppert of Clemson University in South Carolina says that while Uyeda’s work doesn’t prove it will ever be possible to predict earthquakes, it does show the need for more research. “I agree with the authors that the chances of their signals being man-made noise are slim.”

More at: Proceedings of the National Academy of Sciences (vol 99, p 7352)