An Interview With Ray Kurzweil

  • Share
  • Read Later

Techland: For those that don’t make it to the second bridge, what do you believe happens to people when they die?

Kurzweil: Well we can say that people live on in the patterns of information that they leave behind, including the memories of that person among the people who knew them and loved them. Their works and artifacts and ideas.

That’s about all we can say. There’s really no evidence that we live on in any other way that you can point to. I think we’ve rationalized death as a good thing because there’s been no credible alternative up until recently. So we did the next best thing saying, “Oh, that tragedy? That’s really a good thing.” And we’ve invested a lot of effort into these rationalizations.

For example, with my father, I have all of his records. He kept everything. His letters, he was a musician so I have all of his music, photographs, videos, and so on. In the future, AI’s will be intelligent enough to create a virtual person. An avatar in virtual reality that would have some of the characteristics of the person who’s no longer with us based on our memories of that person, all of our files about that person, and their DNA which you could find at their gravesite.

So in that sense we could bring back an avatar that would pass a person-specific Turing test. You could have a healthy debate about what the relation is between that avatar and the original person. My father died in 1970. Had he lived, he’d be much different today than he was then.

But that’s about all we can say.

Techland: What modern day consumer technology impresses you the most?

Kurzweil: I’ve actually been writing an essay on how my predictions that I made in the late 1990s in The Age of Spiritual Machines for 2009 have fared. And 90% of them are correct and some of them became correct very quickly. They looked like they were years away to many people and then a month later, they’re there.

Like high-quality speech recognition on cell phones, or a translating telephone that would convert from one language to another, or augmented reality. Late in 2009, augmented reality applications started appearing on the iPhone. You can point your iPhone at a building and it will recognize the building using image recognition and GPS, and then access a database through wireless communications to essentially allow you to see inside the building. You can see there’s a Starbucks in there and so on.

Recently someone actually put out an application that you can point at a person and there’ll be a little pop up that’ll give you information about that person. I think if we go forward, augmented reality and virtual reality are going to become the next major wave over the next five years or so.

We’ll be putting displays in our glasses, writing images directly to our eyes, creating a virtual display that hovers in the air that can overtake the entire visual field of view and replace it for full immersion virtual reality where we feel like we’re in that environment, not just looking at a screen.

Or augmented reality where you’re in real reality and virtual reality at the same time, particularly where the virtual component is constantly commenting on the real component. We’ll get used to that and wonder how we ever managed without it. There will be constant pop-ups giving you information about people, places, and buildings. We can see the forerunners of that already.

Techland: You’ve invented a ton of stuff over the years. What advice can you offer to potential inventors looking to bring something to market?

Kurzweil: I got into technology forecasting specifically because I was an inventor. And I realized that the key to success as an inventor is timing.

We do some early-stage investing and mentoring, and we get a lot of technology and business plans. I’d say 95% of those teams would build exactly what they say if given the resources but at least 95% of those projects would still fail because the timing is wrong. Not all the enabling factors that are needed for success in the marketplace would be in place at the right time.

Larry Page and Sergey Brin had a great idea about reversing the links on the Internet, and they did it at exactly the right time. The Facebook founders had a great idea about putting these paper Facebooks online and they did that at exactly the right time.

The tools for disruptive change are really in everybody’s hands. In both the Google and Facebook cases, those were done as student dorm projects using $1000 laptops. You don’t have to be a big organization to create powerful, disruptive technologies. And that’s going to be true not just in software, but in areas like biotechnology. The cost of a biotechnology lab is also dramatically coming down.

So the key is timing. You might think the future is unpredictable and, actually, which projects will succeed and fail is unpredictable, but the overall power/price/performance of these fundamental measures of information technology are amazingly predictable.

That was the startling discovery I made 30 years ago. You look at the number of MIPS per dollar that you can buy, it’s an amazingly smooth doubly exponential curve going back to the 1890 census. And nothing had any impact on it, like two world wars, the Cold War, and the Great Depression.

And I’m saying this now not just looking backwards. I’ve been making accurate forward-looking predictions, particularly when it comes to these fundamental measures of information technology.

So the world will be very different in a very short period of time. Think back three years ago. People didn’t use social networks or Wikis or blogs or Twitter. That sounds like ancient history but it wasn’t so long ago. The world will change again just as dramatically in the next three years. People kind of assume that not much is going to change, but that’s really not accurate.

You can predict, quite accurately, certain aspects of the future. In my own case, the fundamental application is still to my own technology projects. For example, we have a project now called Blio, which is an e-reader. It’s a joint venture with the world’s leading book distributor. It’s software that comes with a million free books and is going to be pre-installed on major brands of computers and it’ll be the e-reader for major retail chains.

We have a technology and business plan and, as a discipline, every six months we actually write out what the world will be like in terms of technologies that are important for this type of product. So we will describe exactly what wireless communication will be like, what the platforms will be like, what mobile computing will be like in January 2011, July 2011, January 2012 and so on. We make sure that our plans are appropriate for the world in that description of the relevant technology.

And very few people do that. They just have intuition that things will change a little. People’s intuition about the future is linear, not exponential, which makes a profound difference.

So really consider the future pace of change as a discipline. Be very specific about what the underlying technologies will look like as your project rolls out. Your project has to make sense when it’s finished, not when you start the project. It’s a little like skeet shooting. You have to shoot ahead of the target. Most projects, not just inventions, have a period of time in which they roll out and you have to take this very rapid pace of change into consideration.

Techland: If you could solve one big problem today, what would it be?

Kurzweil: I think the most important problem to solve, because it will give us more time to solve other problems, is reverse engineering biology. To really understand it in information terms. How, exactly, does cancer work?

I’m actually working on that. I’m working on a cancer stem cell project with some scientists at MIT. We believe we’ve actually found the cancer stem cell, so we need to deal with that as well as the cancer cells themselves.

The basic information processes underlying these biological disease processes are not that complicated once we figure them out. And once we figure them out, the remedial solutions readily suggest themselves.

So that’s kind of a race. As we solve these problems, it’ll give all of us more time to solve all the other problems.

  1. 1
  2. 2
  3. 3
  4. Next