7. Brainwaves, Robots and Green Computing

Etech has its odder side. Vacuum packing tofu with herbs, open source 3D printers and questions of whether personal bloggers can ever be anonymous online make their way onto the program because they appeal to the hackers and tinkerers who come to the conference. But sessions also explored how some of what sounds outlandish today could be mainstream in a few years time, like sensor networks and robots.

“The Long Tail” author and Wired Editor Chris Anderson showed how to build your own drone aircraft, using simple hardware (including Lego MindStorms controllers and PDAs). DIY drones can be programmed to fly complex paths using onboard GPS – and can even send back pictures using PDA phone hardware.

Bug Labs brought along their pluggable programmable hardware, which you can use to build and program your own pocket devices. Ambient Devices explored how people relate to devices like glowing orbs that show them what’s going on their own world – and how that relationship could make public transport more convenient by giving you trustworthy information in a personal way.

Meanwhile the team at OpenSpime (inspired by science fiction writer and futurist Bruce Sterling) is planning the architecture of “an Internet of things,” where hardware knows what it is, where it is and how it’s meant to work by using simple sensors to create complex models of the world very quickly. If your car keys were a spime in a network of spimes you’d never lose them, as they’d know they were under a pile of papers on the desk in the den. Or you could get your robot to find them.

Steve Cousins of Willow Garage seeks to develop an open source platform for personal humanoid robots. He wants to create 10 basic models of robots that researchers will be able to improve by using open source techniques. Sharing hardware and control solutions along with new code will accelerate developments in robots, he hopes. In a few years the descendants of these 10 robots might clean your whole house and not just the floor.

Drugs and DNA

Journalist Quinn Norton looked at what counts as medical treatment and what’s unacceptable use of scientific discoveries. When coffee was introduced in the 17th century, the rise of coffee houses – and arguments in coffee houses – led King Charles II to try and ban it. Your cup of java survived and alcohol and tobacco are both legal, but many other drugs are illegal. What should the legal status be for “smart drugs” that enhance memory or cognitive skills for healthy people?

Is choosing to take a drug like Provigil just to stay awake legally or morally different from having it prescribed by your doctor for FDA-approved reasons, such as dealing with sleep disorders caused by working nightshifts? Society is comfortable with pierced ears and cosmetic surgery, less so with tattoos and body piercing. Lasik eye surgery is common but so far only one or two companies have asked employees to have RFID chips implanted for security purposes. Norton had a rare earth magnet implanted in her finger for several months, which meant she could sense electrical fields (as well as pick up paperclips). However, extra senses don’t have to be that intrusive. Brainwave controllers have been used with quadriplegic patients and now they’re being developed as gaming controllers.

Clinical geneticist Hugh Rienhoff is less concerned with smart drugs and is more interested in diagnosing unknown genetic syndromes – like the one his five year old daughter seems to suffer from. When she stayed small and weak for her age despite being a happy and smart child, he remembered some oddities in her appearance at birth and decided he wanted to get her DNA sequenced to help with diagnosis. When he couldn’t find a commercial service to do it for him he used his work skills to get small sections of DNA sequenced in academic labs and compared them against the human genome database.

He was then able to find a mutation and talk to genetic researchers and doctors who were experts on that mutation to find a drug that might help protect his daughter from diseases common to people with that mutation. Sequencing DNA is getting cheaper, according to Pauline Ng who worked on assembling the first full human genome sequence.

The sequence released by the Human Genome Project in 2003 cost $3 billion. In 2007, it cost $1 million to sequence the DNA of one person (who was James Watson, one of the discoverers of DNA), but this year you could do it with $100,000 and a powerful workstation. Two years ago the X Prize Foundation announced a $20-million prize for sequencing the DNA of a 100 people in a week. Today you can get a fraction of your DNA sequenced for $3,000. The implications for personal medicine are excellent and are not just reserved for extreme cases like that of Rienhoff’s daughter. They can determine how cancer patients may respond to different treatment regimes based on genetic predispositions. There are always worrying implications as well: the FBI CODIS database stores the DNA of offenders but DNA analysis is far from perfectly accurate. And what if insurance companies and employers want to test your DNA to put up your premiums or limit liability for work-related illnesses?

Faced with a case alleging it didn’t protect workers from carpal tunnel syndrome, Burlington Northern Santa Fe wanted their workers’ DNA to test for a disease that could have the same effect (although it withdrew the request after a court case).

The ETech attitude is that technology matters everywhere. Stanford law professor Lawrence Lessig, the creator of Creative Commons licenses, suggested that it might be time to turn the principles of technology development onto the political process. "We may be at a time when we need geeks as heroes," Lessig said.

Lessig introduced Change Congress, a Creative Commons-style project for political candidates that can indicate the level of reform they promise, in an effort to switch from lobby funding and earmarks to public contributions. And several of the presentations may have made the audience think twice about upgrading to the latest technology by concentrating on the economics of energy and the difficult of reducing carbon emissions. Squid Labs has already come up with a smart rope that knows how much weight it’s holding up and solar pacing for driveways.

Now Squid Labs co-founder Saul Griffith is running a renewable energy company called Makani. He calculated his personal energy usage in Watts and light bulbs. Counting everything from flights to food to a share of the U.S. Army’s energy usage, the average American uses 12,000 watts a year, which is the equivalent of 120 100-watt light bulbs running 24 hours a day all year long. Bottled water takes 90 watts to manufacture (per bottle), which is the same as a light bulb.

His own usage, including flights to Europe and Australia and cross-state road trips, is closer to 25,000 light bulbs and he plans to cut it significantly by flying less, driving less and eating vegetarian.

Griffith has crunched the numbers on energy use and carbon emission but he couldn’t be accurate about where CO2 really comes from because we don’t have enough sensors to find out. We also cannot accurately assess the impact of CO2 emission on the climate because we don’t have a powerful enough computer to model the entire climate. That would take a zetaflop computer the size of Paris, using a terawatt of power, HP Senior Fellow R. Stanley Williams, said. If Williams is correct, then that would mean modeling the climate would have more impact on the climate than anything else we do.

The human brain is the equivalent of 100-million gigaflops using only 10 Watts of power (and a gigaflop is billion floating point calculations per second); that’s 100 petaflops and we’re only just on the verge of petaflop computing. Exaflop computing is at least 10 years away and going faster is going to require a different way of computing, using nanoscale materials. Williams predicts chips will go optical, with photonics and nanoscale switches hybridized onto silicon to give faster performance in smaller sizes and at lower power. Williams envisions an optical computer with individual nanowires that act like wave guides to form a switching network built up as an interconnected grid. No longer a binary computer, photonics-based electronics is more like an old-fashioned telephone network – or the synapses in the human brain.