This article was originally published in the Summer 2002 issue of LINK magazine.
Linking your computers by tangling wires through walls and under carpets is already a bit quaint—very twentieth century. Yet only a few years ago, an affordable wireless home network seemed decades away. (For that matter, so did a computer mouse with no rolling ball.) Which PC technologies are way out there today, but might be everywhere tomorrow?
At Caltech, there's a monkey who can do something with a computer that you can't: move a cursor on screen just by thinking about it. Under the sea, U.S. Navy submarines send top-secret communications in nanosecond bursts. At a lab in Menlo Park, California, researchers are figuring out how to make computer circuits that come from a printer, not a clean-room assembly line.
Here's a look at these and a few other technologies that might take this decade's personal computing experience to the next level.
On Valentine's Day 2002, the U.S. Federal Communications Commission permitted commercial use of ultrawideband (UWB) communications for the first time. (Find out more at www.uwb.org.) But that ruling comes with some strong power and radio wavelength restrictions, because UWB is radically different from other wireless data standards. It avoids specific narrow radio bands, instead cutting right across huge swaths of radio frequencies already used for other purposes, and layering its own signals on top by using extremely short (billionths of a second), low-power, precisely timed bursts to transmit information.
One benefit is that many different UWB transmissions can coexist with each other, and with more traditional wireless signals. UWB also allows remarkable range and throughput using very low-powered and inexpensive equipment: a wireless LAN running at up to a gigabit per second (a hundred times faster than current wireless networks) needs a tiny fraction of the power of a cellular phone. UWB devices can determine their physical location within 10 centimetres, and the precise timing of transmissions makes them extremely secure—enough that the military developed UWB in the 1960s to communicate with distant submarines.
The technology is ready. UWB computer networking equipment will likely be shipping later in 2002. But there are downsides too. Law enforcement and firefighters complain that the FCC's restrictions make UWB too low-power to "see" through walls—something it could do at higher powers, but which could have privacy implications in a wider commercial context. On the other hand, the military, users of the Global Positioning System (GPS), and radio astronomers are concerned that widespread UWB could interfere with their particularly sensitive conventional radio systems. That's because, to a regular radio, UWB is just noise—and a lot of UWB transmitters could raise the overall background radio noise level noticeably.
Chances are that UWB will come, maybe this year, maybe a few years down the road. Then you can take all your spare network wires and build a hammock in the yard, where you can use your UWB-equipped laptop in the sun. If you're lucky, that laptop might be as thin as a piece of paper too.
In Cambridge, Massachusetts at the beginning of 2002, a company called E Ink (eink.com) produced a thin, flexible prototype silicon display that you can bend, twist, roll up, or drop without damage. In 2001, Lucent Technologies (lucent.com) created a similar display using flexible plastic transistors. (Just in time too—Stanley Kubrick's 1968 film 2001: A Space Odyssey expected that astronauts would read newspapers and watch TV on electronic ink sheets by now.)
One firm, Rolltronics (rolltronics.com), seeks to mass-produce its flexible circuit sheets using sophisticated printing techniques, instead of in expensive factories with photolithographic clean rooms and technicians in sterile bunny suits. Combined with flexible electronic ink, thin-film batteries, and memory (also in development), a few years of research should yield a computer you can roll up and put in your pocket or knapsack.
But there's a problem: keyboards and mice don't take well to being rolled up. That's where our next technology comes in.
Profoundly disabled people have long been frustrated by the limitations of input technology. Think of the genius brain of Stephen Hawking, whose body can only make rudimentary movements that he must control to make a computer speak for him. What if the computer could read his mind instead? And what about the rest of us? The keyboard as we know it is over a century old, and even the computer mouse is approaching its fortieth birthday. Isn't there something new out there?
At the California Institute of Technology, Daniella Meeker and her research team are teaching a computer to read a monkey's mind. They have determined that a small group of cells in the posterior parietal cortex (a part of the brain shared by all primates, including humans) initiates specific movements in the body. By monitoring what these cells did when the monkey played a simple video game, then training the monkey to think about playing the game without moving, the team had a computer read the monkey's brain signals with electrodes. Eventually, the monkey was moving a cursor around the screen by thought alone.
There are two problems with bringing mind-reading computers to humans. First, interpreting complex tasks (such as typing a letter) from brain waves is still some years away. Second, few people find keyboards annoying enough to undergo brain surgery to get rid of them. Still, scientists expect more sophisticated mind reading to be possible, and others are working on ways of tracking brain signals without surgery.
One other difficulty arises from the Caltech experiment. Once the monkey got used to playing the game by mind control, it didn't want to use its hands to play anymore. If today's couch potatoes seem lethargic, imagine a generation that doesn't even need to reach for the remote. You can find out more by thought-controlling your Web browser to www.brainland.com and checking out the "Basic Research" links.
Even if they might read our minds, computers still aren't very smart. They don't think, they compute, and they do it very quickly. Most of the time, our computers are waiting for us or sitting idle while we sleep, eat, and do all those other things that make being alive so inconvenient. In recent years, a number of organizations have harnessed that idle time. The earliest offered challenges for testing encryption codes. Then came SETI@home (setiathome.berkeley.edu), which created a global super-supercomputer to analyze radio telescope data for signs of alien intelligence (no luck so far). More recent efforts tackle protein folding and other biotechnology issues.
What if all these schemes and many others could come together? Imagine if there were an infrastructure where all Internet-connected computers could share their spare time to perform weather simulations, genetic analysis, and special-effects rendering-or even back up one another's files so that no faulty hard drive could ever lose data permanently again. (Everything would be encrypted, of course, so you couldn't grab a mini-chunk of a scene from the new Star Wars movie, even if your computer was building frames for it.) Then imagine a micropayment scheme that transfers tiny amounts of money from one person or organization to another according to how much computing power they make available. It would be a whole new economy.
Those who advocate the idea call it an Internet-scale operating system (ISOS). Current prototypes in the scientific community are called grids, and include the Eurogrid (eurogrid.org), the Grid Physics Network (griphyn.org), and the Particle Physics Data Grid (ppdg.net). Scientific American magazine (sciam.com) published "The Worldwide Computer" on the subject in its March 2002 issue.
Okay, we've been way out there now for a few paragraphs. How about something more immediate and tangible, like a cheaper digital camera that takes way better pictures?
A California start-up company called Foveon (foveon.com) has developed a digital photo sensor it calls the X3. Unlike those used in digital cameras today, the X3 uses a multi-layer approach that determines the colour of incoming light by how deeply it penetrates the sensor surface. One bonus of the design is that a 3.3 megapixel X3 sensor produces images with the quality of a 7 megapixel traditional digicam-and it's cheaper, smaller, and faster. Sigma's (sigmaphoto.com) high-end $3000 (US) SD9 is the first X3-equipped digital camera, but less expensive models will certainly arrive soon.
So in a few years, when you take your high-quality digital photograph and use thought control to transfer it over an ultrawideband network to the rolled up computer in your pocket, which is working on the latest analysis of the human genome with tens of thousands of others around the world, remember that you read about it here first.
If you found this article useful, feel free to consider making a donation (any amount, credit cards accepted), which helps pay for hosting this website. Thanks!
Page BBEdited on 20-Mar-04 (originally published July 2002)