Followers

Friday, June 6, 2008

The Future Is Now? Pretty Soon, at Least

Before we get to Ray Kurzweil’s plan for upgrading the “suboptimal software” in your brain, let me pass on some of the cheery news he brought to the World Science Festival last week in New York.

Do you have trouble sticking to a diet? Have patience. Within 10 years, Dr. Kurzweil explained, there will be a drug that lets you eat whatever you want without gaining weight.

Worried about greenhouse gas emissions? Have faith. Solar power may look terribly uneconomical at the moment, but with the exponential progress being made in nanoengineering, Dr. Kurzweil calculates that it’ll be cost-competitive with fossil fuels in just five years, and that within 20 years all our energy will come from clean sources.

Are you depressed by the prospect of dying? Well, if you can hang on another 15 years, your life expectancy will keep rising every year faster than you’re aging. And then, before the century is even half over, you can be around for the Singularity, that revolutionary transition when humans and/or machines start evolving into immortal beings with ever-improving software.

At least that’s Dr. Kurzweil’s calculation. It may sound too good to be true, but even his critics acknowledge he’s not your ordinary sci-fi fantasist. He is a futurist with a track record and enough credibility for the National Academy of Engineering to publish his sunny forecast for solar energy.

He makes his predictions using what he calls the Law of Accelerating Returns, a concept he illustrated at the festival with a history of his own inventions for the blind. In 1976, when he pioneered a device that could scan books and read them aloud, it was the size of a washing machine.

Two decades ago he predicted that “early in the 21st century” blind people would be able to read anything anywhere using a handheld device. In 2002 he narrowed the arrival date to 2008. On Thursday night at the festival, he pulled out a new gadget the size of a cellphone, and when he pointed it at the brochure for the science festival, it had no trouble reading the text aloud.

This invention, Dr. Kurzweil said, was no harder to anticipate than some of the predictions he made in the late 1980s, like the explosive growth of the Internet in the 1990s and a computer chess champion by 1998. (He was off by a year — Deep Blue’s chess victory came in 1997.)

“Certain aspects of technology follow amazingly predictable trajectories,” he said, and showed a graph of computing power starting with the first electromechanical machines more than a century ago. At first the machines’ power doubled every three years; then in midcentury the doubling came every two years (the rate that inspired Moore’s Law); now it takes only about a year.

Dr. Kurzweil has other graphs showing a century of exponential growth in the number of patents issued, the spread of telephones, the money spent on education. One graph of technological changes goes back millions of years, starting with stone tools and accelerating through the development of agriculture, writing, the Industrial Revolution and computers. (For details, see nytimes.com/tierneylab.)

Now, he sees biology, medicine, energy and other fields being revolutionized by information technology. His graphs already show the beginning of exponential progress in nanotechnology, in the ease of gene sequencing, in the resolution of brain scans. With these new tools, he says, by the 2020s we’ll be adding computers to our brains and building machines as smart as ourselves.

This serene confidence is not shared by neuroscientists like Vilayanur S. Ramachandran, who discussed future brains with Dr. Kurzweil at the festival. It might be possible to create a thinking, empathetic machine, Dr. Ramachandran said, but it might prove too difficult to reverse-engineer the brain’s circuitry because it evolved so haphazardly.

“My colleague Francis Crick used to say that God is a hacker, not an engineer,” Dr. Ramachandran said. “You can do reverse engineering, but you can’t do reverse hacking.”

Dr. Kurzweil’s predictions come under intense scrutiny in the engineering magazine IEEE Spectrum, which devotes its current issue to the Singularity. Some of the experts writing in the issue endorse Dr. Kurzweil’s belief that conscious, intelligent beings can be created, but most think it will take more than a few decades.

He is accustomed to this sort of pessimism and readily acknowledges how complicated the brain is. But if experts in neurology and artificial intelligence (or solar energy or medicine) don’t buy his optimistic predictions, he says, that’s because exponential upward curves are so deceptively gradual at first.

“Scientists imagine they’ll keep working at the present pace,” he told me after his speech. “They make linear extrapolations from the past. When it took years to sequence the first 1 percent of the human genome, they worried they’d never finish, but they were right on schedule for an exponential curve. If you reach 1 percent and keep doubling your growth every year, you’ll hit 100 percent in just seven years.”

Dr. Kurzweil is so confident in these curves that he has made a $10,000 bet with Mitch Kapor, the creator of Lotus software. By 2029, Dr. Kurzweil wagers, a computer will pass the Turing Test by carrying on a conversation that is indistinguishable from a human’s.

I’m not as confident those graphs are going to hold up for fields besides computer science, so I’d be leery of betting on a date. But if I had to take sides in the 2029 wager, I’d put my money on Dr. Kurzweil. He could be right once again about a revolution coming sooner than expected. And I’d hate to bet against the chance to be around for this one.

Original here

No comments: