Ray Kurzweil: Don't fear the nanofuture

Among other achievements, Kurzweil, who has worked on artificial intelligence and pattern recognition technologies for over four decades, invented the first reading machine for the blind.

In his latest venture, the tech entrepreneur has essentially taken on a Johnny Appleseed role, demonstrating how breakthrough--even sci-fi sounding--technologies are going to improve the way people work.

In books and speeches, Kurzweil has sketched out a scenario where the distinction between human and machine blurs, an emerging era where microscopic "nanobots" will be able to trigger images allowing people to operate in virtual reality whenever the mood or need strikes.

But Kurzweil's musings have also triggered a larger debate about what it means to be human in a future marked by a fusion with robotic technologies and synthetic personalities. Most notably, he has been paired as the sunny optimist vs. Bill Joy, the chief scientist at Sun Microsystems, who last year penned a long article that raised a red flag about some of Kurzweil's ideas.

The inventor-author nonetheless is optimistic about the future and last month gave a public demonstration of how a singer could be transformed into a virtual reality persona in real time. Kurzweil talked about his ideas in a recent interview.

Where do you go in terms of bringing the concept of conversational Web site avatars into more regular, everyday use?
We're investigating options for Web-based speech recognition. This will take time, but eventually you will be able to speak to her and get good quality back. It'll probably take six to 12 months. We are doing constant work on updating her knowledge base, but she also has a lot of general knowledge from conversation. Of course, she doesn't yet pass the Turing test. (The Turing test was developed by computer scientist Alan Turing to determine whether computers can "think.")

Take the 30,000-foot view. What do you envision as the predominant benefit society will derive from virtual reality?
There are several ways to approach this. I think avatars will be a near-term trend over the next several years. We'll see them as a very useful human interface. They're entertaining but also a good way to interact with machines.

For example?
Well, I don't like when forms pop up (on a Web page). It's a struggle. I would rather say what I want them to do. We'll interact with virtual personalities, virtual sales agents, virtual information assistants--connected with search engines which will guide our interactions with machines. So you don't need to know anything other than what you want to accomplish from the machines. You'll quickly adapt for the interaction to be successful, just as you do when interacting with people.

What about in terms of virtual reality?
In terms of virtual reality, I see that as a profound transformation for our civilization. We've already had one form of virtual reality: auditory; when we use the phone, it's just as if you and I are together. The next step is to add the visual sense. That will emerge gradually but should be here by the end of decade, with a quite compelling, full auditory, shared environment. The electronics for this will be entirely unobtrusive, woven into our clothing and eyeglasses and completely invisible. We'll have detection of our body movements and high-bandwidth, wireless connections to the Internet so that walking around, you'll have things in your visual field and be able to connect to Web and its virtual manifestations.

What comes after that?
If you go out 30 years, virtual reality moves inside our brain with the nanobot scenario. Because of the shrinking of technology, it's a conservative scenario to say we'll be able to send cell-sized robots into our bloodstream, through capillaries of our brain. They'll be communicating with each on a wireless area network, and there'll be communications with your biological neurons. If you want to be in real reality, the nanobots will do nothing; if you want to go virtual, they'll replace signals you would have been receiving in the real environment.

Have you heard much from skeptics asking why you would think that many people ought to prefer full immersion into virtual reality over real reality?
Sometimes people say that. But there is confusion over the word "virtual." People had the same reaction to the telephone...It's very easy now to add the visual sense; we do it crudely with videoconferencing.

What are the implications for the way we work?
You can virtually do all business meetings. This would replace the bulk of business meetings that take place. It's going to be real communication between real people.

You've said that by 2009, we'll have ubiquitous, full-immersion, shared visual-auditory virtual reality environments. In practical terms, what need be the sequence of events to realize that future? And which industry--or industries--do you think will take the lead?
All these things require certain enabling technologies. For instance, there are a number of vendors today who offer retinal projection glasses. They don't yet have the resolution or look completely normal, but it's not 2009 yet. We already have wearable computing. But we're not yet at the point in Moore's Law where it's become completely invisible. We need to get personal local area networks. And certainly, we will have systems like that so we don't have to carry all those wires. So your personal computer will turn up in your clothing and your display will be in your glasses. By 2009, we will have very high bandwidth, and the electronics for everything will be so small that it will be embedded everywhere.

OK, let's move to a different subject. I read in an essay where you rejected Bill Joy's prescription to relinquish what he called "our pursuit of certain kinds of knowledge." Seems that you're now being twinned with him as the sunny technologist vs. the gloomy technologist. Anyway, what do you think are the implications for technology following Joy's dictum in its totality?
I don't think you can follow Joy's dictum. It reflects an unrealistic concept of technology in our society. Dangerous technologies are the same technologies giving us the benefits. Nanotech is not one field; it's the inevitable end result of miniaturization. That inevitably gets you to nanometer scales in 25 years. There's no way you can stop it.

Is Bill Joy going to have Sun Microsystems stop its pursuit of faster computers and more effective network protocols.? It's an utterly unrealistic view. (Unabomber) Ted Kaczynski makes this very same point. He says that he acknowledges the profound benefit to technology. The question comes up: Why not keep good ones and get rid of bad ones? It's the same technology.

Does he have a point expressing skepticism about our ability to do the right thing with technologies, which may have potential to become double-edged swords? Cloning research comes immediately to mind. Or the controversy over the work done in creating the atomic bomb?
There's absolutely a double-edged sword. Once you get past the feasibility issue, it becomes a desirability issue. My prescription is lots of vigilance by responsible practitioners, law enforcement where necessary, building safeguards into them and building environmental immune systems. The dangers are very profound.

You have talked about the intertwined promise and peril of this century in the context of technological change. Can you expand on that?
If you look at today, you can see tremendous benefit from technology by looking back 200 years when human life expectancy was half what it is today. We're measurably better off, but there are also many dangers. We have the firepower to blow life off the planet, and there are lots of other weapons of mass destruction. Technology is powerful and gives great power to individuals. It's very democratizing but also enables destructive elements. We're talking about the duality of human nature, and it can be very creative or also very destructive.

How close are we from the day when machines rely not just on pattern recognition but can actually make the leap to possessing imagination?
Imagination and creativity--some of the higher-order emotional responses like humor, love, joy--this is the cutting edge of human intelligence. That's exactly what machines will achieve when we emulate human intelligence. My sense of that is the year 2030. We will have a fully reverse-engineered human brain and understand how intelligence works. We can already show isolated examples today. We're not there yet, but you can already see some sparks.

You've participated in the birth of the PC era and have witnessed much. What's been the biggest surprise in the way artificial intelligence and pattern recognition technologies have turned out?
I've been in computers since 1960 and have gained a deepening appreciation for what's really different about human intelligence. And it's the opposite of what you might expect: It turns out to be easy to replicate most adult achievements, like playing chess or guiding cruse missiles, etc. At least a lot of them we can replicate in machines. Every time a machine does that, it's the nature of artificial intelligence to say, "Well, that wasn't very difficult after all." The hard things are what children do, recognizing parents, or playing with a dog.