Q&A: Futurist Ray Kurzweil on Google, reverse engineering the brain and more
Wed, Jan 23, 2013 (6:34 p.m.)
I know you were recently snatched up by Google. What are you working on at the moment?
I’ll be working on enabling computers to understand natural language. When you write a news article or a blog post you’re not just creating a bag of words, you’re creating semantic meaning … It’s remarkable how search works, but it could be even more powerful if it understood the key questions you had on a detailed, semantic level, and if it actually read and understood, at least on some level, all these billions of web pages and book pages out there. So that’s the goal, as well as getting to know the user, knowing what the user’s own concerns are and what kind of information you’re looking for by actually understanding natural language.
Your book is about the reverse engineering of the brain. What does that mean?
Reverse engineering simply means to understand, but to understand at detailed level, sufficient that we could actually recreate its methods. There are three benefits to reverse engineering the human brain. One is to do a better job of fixing it. So we have certain drugs, let’s say an anti-depressant, that basically treats the human brain as a chemical stew, when in fact it’s a very elaborate network. We’re beginning to be able to have interventions that really do treat the brain as a very delicate circuit. We have neural implants for Parkinson’s disease and cochlear implants for the deaf that are very isolated and selective in the part of the brain that they’re interacting with. That’s going to become much more pervasive. Ultimately we can send computers into the brain without surgery by sending them in through the bloodstream.
Secondly, we can create much more intelligent computers if we can learn the secrets of how the human brain creates intelligence. Thirdly, it gives us more insight into ourselves. We can’t really understand who we are unless we understand how our brains work, because it’s such a vital part of who we are.
- Ray Kurzweil
- January 27, 7:30 p.m., $24-$59
- Smith Center's Reynolds Hall, 749-2000
Can you describe your theory of technological singularity?
Well, I began to be interested in how technology evolves about 30 years ago because I realized that your invention has to make sense for the world that will exist in the future when you finish your project. If your project takes, let’s say three years, it’ll be a very different world. And if you need any evidence of that, just look at how different the world is today than it was three years ago, when most people didn’t use social networks or wikis or blogs. It gives you some idea of how quickly it’s changing, and from that study I came away with a thesis called the Law of Accelerating Returns, which basically says that information technology is progressing exponentially, not linearly. It basically doubles its power every year approximately. Our intuition about the future is that it grows in a linear manner, as in 1, 2, 3, 4, 5, and at step 30 you’re at 30. An exponential progression is 2, 4, 8, 16, and at step 30 you’re at a billion. It makes a huge difference.
We’ve already seen that kind of improvement from this exponential growth of information technology. And it applies to every form of information technology, including now, biology, which we are now understanding as an information process. Communication technologies are trillions of times more powerful than they were a century ago, when we were sending Morse code over AM radio, for example. If you follow the inevitable trajectory of ongoing progress at an exponential pace, we get to a point where we will merge with this technology and greatly expand our own intelligence.
With regards to the 2040s, we’ll have expanded our intelligence a billion fold by merging with the nonbiological intelligence we’re creating. We’re already on that path. We are smarter today than we were 10 years ago, say before we had Wikipedia, or 15 years ago, before we had Google. These tools do make us smarter. I’ve been managing work groups for 45 years and I can have a group of a few people now and in a few weeks accomplish what used to take hundreds of people years. We’re far more effective and productive as a result of these brain extenders today. Go out another 30 to 35 years and we’ll have expanded our intelligence a billion fold. And that’s such a profound change that we borrow this metaphor from physics and call it a singularity, basically a singular change coming in human history.
You’ve predicted the singularity will occur in 2045. What are your thoughts on what the world will be like when that happens?
Well, we’ll see many differences long before that. As I mentioned, we’re treating health and medicine now as an information technology. So already we’re re-growing organs made from your own stem cells. It’s been successfully done in animals, and it’s actually been successfully done in humans for simple organs, and there [are] scientists ready to take the next step to complex organs. I’m involved with a company where we’re re-growing lungs for people with profound lung diseases.
We’re going to be able to rejuvenate every organ in our bodies from the inside. We’ll be able to have nanobots, little blood cell-sized devices that augment our immune system that are more intelligent than our white blood cells. Our immune system naturally is very helpful—we wouldn’t last long without it—but it’s also very limited. It doesn’t recognize cancer, for example. It thinks that’s you. Sometimes it thinks that you are the enemy, and those are autoimmune disorders. We could create an extension to the immune system that fixes these limitations. The immune system is pretty useless against most viruses, for example. We’ll be spending … most of our time in virtual reality environments created from inside the nervous system. We’ll have computers inside the brain, kind of a la The Matrix, but these ones necessarily won’t be dystopian ones. We will have full-emergent virtual reality, incorporating all of the senses, and you can have a different body in these virtual environments. So these are just a few of the implications.
This will probably sound silly, but talking about robots has me thinking about The Jetsons. Will we soon depend on robots to do, well, everything?
We will have robots. There are robots in the world now. They have a limited intelligence but the intelligence of everything is going to get greater and greater. You go out to the 2030s, I think that the AIs will be as intelligent as humans, and I talk about the philosophical, social and political implications of that in my book. I don’t see it so much as a conflict between humans on one side of the room and machines on the other; I think it’s going to be all mixed up. Basically, we’re going to use these intelligent machines to make ourselves smarter, and as I argued earlier, we’re already doing that. We will be able to do that with even more intelligent machines as we go forward, but there will be robots to do chores.
You know, robots already do most of the work in factories. In fact, a third of the population worked in factories in the 1900s, [and] toady it’s 2 percent. Another third of the population worked on farms, and it’s also been mechanized with robotic devices and it’s down to 2 percent. So we’re going to continue in that vein. But that actually doesn’t mean that employment is less, employment is actually higher. We have a higher percentage of the population working than we did 100 years ago, and those new jobs pay 10 times as much … because they’re at a higher intellectual level and we actually use our technology to enable us to do work at a higher intellectual level.
Will robots be capable of emotion?
I think they will have emotional intelligence. They’ll be able to exhibit emotions, which will be convincing … They’ll be indistinguishable from human intelligence, and in order to do that it has to be as emotional as a human. It has to be able to get the joke or be funny or be sexy or be loving. These are not sort of sideshows to human intelligence, it’s actually the most impressive thing we do, is to exhibit these kinds of emotional reactions. So in fact, that is the way in which humans excel right now compared to computers. If you were to say, “No, no, intelligence is really logical thinking, not this emotional stuff, that’s a distraction,” then computers are already smarter than we are. But in fact, our emotional intelligence is a very intelligent type of behavior and a very effective one. But that’s something computers will master by 2029.
What will you discuss during your speaking engagement at the Smith Center?
Well, really, all of the things we have been talking about. The exponential growth of information technology, how it’s going to transform every aspect of our lives, how the human brain works, how we’re mastering those methods and that we’ll create artificial intelligence in our own image, [and] how that will improve human life, what some of the dangers are, [and] how we can deal with those.