Fast Company: First off, without death, CEOs will never give up their jobs. There won’t be any succession plans.
Ray Kurzweil: I don’t think we need to kill people off to provide opportunity for new leadership and creativity. The marketplace of ideas and technologies is going to expand -- it has been for years. Look at the computer industry. 60 years ago it was a handful of research projects, and now it’s a trillion-dollar industry.
FC: But biotech? Who’s to say how quickly it will advance?
Kurzweil: A lot of people say you can’t really tell the future, and there are certain things that are hard to predict. What will Google’s stock be three years from now? That’s hard to predict. But if you ask me what it will cost to sequence a base pair of DNA in 2010 or the cost to move a megabyte of data wirelessly in 2015, those things turn out to be remarkably predictable.
FC: To the point that we can program our own biology?
Kurzweil: Information technology is affecting almost every field. We’re now understanding biology as information processing; we’re learning to understand the processes underlying these biological pathways. Whereas drug discovery used to be literally that, discovering drugs, which is to say finding something that happened to work. Now we’re entering an area where we actually understand the exact sequence of biological events. We can intervene very precisely by blocking one key enzyme, one key step.
FC: Alright, before we go any further, tell me about your new book.
Kurzweil: It’s an urgent message to my baby boomer peers, 99% of whom are oblivious to this perspective. People have a very conventional sense of the cycle of life. They just don’t have a sense that they could master the biology that’s controlling their progression towards disease and aging. We’re well along in understanding and reverse engineering the dozen or so biological processes that describe aging. It’s not too late for baby boomers to reverse those processes. We have the tools right now to slow down aging sufficiently so that most of us can remain in good shape until we do have the tools. My view is that I’m reprogramming my biochemistry the way I’d reprogram my computers. I’m measuring 60 different levels on a regular basis. It’s definitely working; I would have heart disease otherwise.The common wisdom is that health is 80% genetics and 20% lifestyle. And that’s true if you take the conventional watered-down approach. If you’re aggressive, you can overcome, well, I wouldn’t say anything -- but almost anything.
FC: Would you want to be immortal if the opportunity presented itself?
Kurzweil: Well, I think the opportunity is presenting itself. Our mortality is something that should be in our hands; it’s something I want in my hands. Science and technology are accelerating. I believe we’ll demonstrate a mouse that doesn’t age within approximately a decade. And within a decade of that we’ll translate that into human therapies.
FC: Birth and death are nice bookends. How will removing them change our philosophy of life?
Kurzweil: It’s already changing. There are lots of people in their 60s who look sexy, who are intellectually vital, who continue to contribute. We’re in an era where people’s contributions are primarily intellectual. Later in life, people have accumulated experience and wisdom, so they’re very much in a position to contribute to society. Our perception of someone 65 earlier in the century was very different. Life expectancy was 37 in 1800. It was 55 in 1900. Now it’s pushing 80. We continually push that back and I think it’s going to change very quickly as we get more powerful tools.We can talk poetically about how aging is natural, but the reality is if you visit an old-age home you see people who’ve lost their loved ones and have lost their faculties. It’s really a tragic situation and it’s not something I desire. I want to keep my faculties.
FC: What about the costs?How is society going to support the cost of all these people?
Kurzweil: We’ll be creating a great deal of wealth. Not just in dollars, but also in what a dollar can buy. We have 50% deflation in information technology; you can buy the same digital camera today for half the price it was 12 months ago. Same specs. Go out to 2020, what you can buy in terms of information for a dollar will be quite vast. With nanotechnology we’ll be turning information into a wide range of products -- including food -- with very inexpensive materials.
FC: When will people choose to stop aging?
Kurzweil: The killer app for nanotechnology is nanobots. Some will be in the environment, cleaning up, providing energy. Some will be involved in automated manufacturing. Some will be in our bodies, keeping us healthy from the inside. Destroying pathogens, getting rid of toxins, killing cancer cells. It will be routine.
Kurzweil: Making sure our bloodstream has all the nutrients it needs regardless of what we eat. We’ll ultimately disconnect the sensual and social pleasures of eating from the biochemical task of keeping an optimum set of nutrients in our bloodstream.We’ve already separated the biological purpose of sex from the social and sensual aspects of it. We’ll discover the opportunity to express ourselves in different ways. We’ll have the opportunity to be different people. It’s one of the features of virtual reality. I demonstrated this with contemporary technology at TED 2001. I went on stage. I was wearing sensors under my clothing. Computers picked up on my movement, and on a big 12-foot screen in real time it created a pretty life-like animation of a young woman, Ramona. And my voice was transformed to a woman’s voice. It looked like Ramona was giving the demonstration. I sang "White Rabbit". Then my 14-year-old daughter got up, and her body was transformed to Richard Saul Wurman. Warner Bros. heard about this and Al Pacino winds up doing exactly what I did in Samone. We have technologies today that do that to a limited extent. Makeup. Fashion. I actually found it to be a profound experience. Once we put on the sensors and got all the equipment going, I was doing this and looking at myself in the cyber mirror. Instead of seeing what I usually see in the mirror I saw myself as someone else. It was a liberating experience. And we do have these other personalities in ourselves that we’d like to express. Couples could turn themselves into the other and get some sense of what it’s like to be someone else. A lot of the misunderstanding in the world is that we don’t see ourselves in other people’s shoes. Personal relationships. Education. A student could actually be Ben Franklin in a virtual constitutional conference instead of just dressing up. A lot of psychological exploration. Certainly entertainment and games. I think we’ll realize that we like expressing ourselves in different ways. We do that to a limited extent. We dress up differently. We put on a tuxedo for one event or blue jeans for another event. We’re changing our appearance a little bit. But we’ll be able to go beyond clothing and fashion and hairstyle and makeup.
FC: Will people begin to develop distributed intelligence?
Kurzweil: One of the ways in which our biological intelligence is limited is that we have only limited ways of hooking up our intelligence to others. We have some ways. We have the Internet, language, books, magazines. We have been able to pool human intelligence across individuals. Now that we have computers and the Internet to gather our knowledge and allow us to search through our knowledge, our ability to do intellectual achievements has grown. But we still can’t hook up the resources of one brain to another. Computers can do that. You can take a network of 10,000 computers and they can create one supercomputer and very quickly share knowledge and data and have all the different processes working on the same problem. Then they can be made separate again.That’s one of the profound benefits of non-biological intelligence, that it can pool its intelligence. I call that falling in love. Human beings, we can kind of merge our thinking with another person. We call that falling in love. But our ability to do that is subtle and fleeting and not something we can control very easily, whereas machines can do that very easily. My view is we will develop a non-biological component of our thinking as we begin to introduce non-biological processes into our brains. We’re in the very early stages of that today. There are people walking around in New York who are cyborgs. They have computers in their brains. For example, the FDA-approved implant for Parkinson’s disease. These are people whose portion of the brain was destroyed by this disease. They have this implant. The implant actually does what those biological neurons used to do. The neurons that are nearby are getting signals from the electronic device just as they used to get signals from the biological neurons that were working. And they’re perfectly happy to get the signals from the electronic device. This hybrid of biological and non-biological components works perfectly well. And in fact the latest generation of this particular implant allows downloadable software from outside the patient. So you can actually download a software upgrade from your neural implant from outside. This is a very early stage -- and it requires surgery -- but ultimately we’ll be sending nanobots, which will have computational resources and communication. We’ll send billions of them through the capillaries of the brain. They’ll be able to communicate wirelessly, non-invasively with our biological neurons. If you go out to 2030, say, and talk to a person of biological origin, they’re going to have a lot of non-biological processes running in their brain. As you interact with them, you’ll be interacting with someone who’s a hybrid of non-biological and biological intelligence. We know that biological intelligence is pretty fixed in its architecture. Today, we have approximately 10^26 calculations per second in the humans species. 50 years from now, the power of our biological thinking will still be 10^26 power. It’s not going to grow. Non-biological intelligence basically doubles every year. The crossover point will be in 2020s. You get to the 2030s and 2040s, the non-biological portion of our thinking is going to be millions of times more powerful than the biological portion. So if you talk to a person of biological origin, the fast majority of their interacting is going to be non-biological.
FC: You say "a person of biological origin" almost as if to imply that there might be people of non-biological origin?
Kurzweil: If the biological portion is becoming fairly insignificant, some people won’t necessarily have one. And we’ll have AI operating at human levels. My position is that by 2029 computers will pass the Turing test, which is to say they’ll be indistinguishable from biological intelligence. But its not going to be a clear distinction, because there’s going to be biological people with non-biological processes running in their brain, there’s going to be non-biological computers that act human because they’re based on the reverse engineering of the human brain. Even that in my view is derivative of human intelligence. It’s the expression of the human civilization. These are not intelligent machines coming from outer space, invading the planet. It’s emerging from within our human civilization. Civilization is already a biological / non-biological hybrid. We do fantastic things that would be impossible without our technology. It’s the technological portion that’s exploding exponentially. We’ll have human-like intelligences that don’t have a biological substrate.
FC: How would distributed technology be manifested? Teams upload and download information with each other?
Kurzweil: We have very efficient ways of sharing information between our personal computers now. Our personal computers are outside our bodies and brains, but just barely. I talked to a woman recently who said her son’s personal computer may as well be inside his brain because it’s an extension of him, and he carries it everywhere he goes. When she comes in the room, she’s just another window because he’s got six windows open on his screen and she’s standing there in the doorway, which is another window. And he’s timesharing between her and the other windows. By early in the next decade we won’t be carrying around these physical objects. Images will be written directly onto our retinas from our eyeglasses or contact lenses and the electronics will be woven into our clothing. We’ll have very high-speed wireless connection at all times. And then it will make its way inside our bodies and brains. It will be a very gradual, incremental process. The way we share information now very fluidly between our personal computers will obviously also happen when these computer processes are running inside our brains. It will be very fluid.
FC: So when will we start seeing something like this?
Kurzweil: But it's affected already. I have people around the country, and it's only a subtle difference between working in our office and people who aren’t. It used to be a big challenge. It wasn’t that long ago. A little over a decade there was no Web. You know what year the first reference to the phrase "World Wide Web" appeared in the New York Times? 1993. Even early adopters didn’t get involved in email until 1994, 1995. I’ve been in business for a few decades. In the early 1990s it was very hard to have someone working with you who wasn’t in your office. Now that’s very routine and we have very powerful ways to share all kinds of knowledge. And when we have really ubiquitous high-quality audio-visual virtual reality, which I think is coming soon, that will be another major step in the ability to work together no matter where you are. And once we have these processes running inside our bodies and brains, which is a couple decades, that will be another major step in the ability to work together and the intimacy of that.
FC: We have five senses for uploading information to the brain. What you’re talking about would introduce a sixth, where knowledge set aside in a storage device can interface with our brain. We’ll have to make the two processes at the same biological pace.
Kurzweil: You’re bringing up an important issue, which is the speed of the process: biological vs. non-biological. Of course, this analogy comes up a lot when you talk about little companies merging with big companies and the big company has these big gearboxes with big gears that move very slowly and the little company has these small quickly moving gears. Sometimes those gears strip when they try to merge. So it's an analogy in terms of large and small organizations integrating. We have a similar issue when we try to marry biological and non-biological thinking, which ultimately will be a lot faster. Our biological thinking takes place at chemical switching speeds, which are a few hundred feet per second. And that’s the speed with which the chemical gradient moves along the axon. They take place at these very slow speeds and the reset time is about five milliseconds. In a typical dentrite that’s about 200 transactions per second. These are very slow speeds. Electronics are about a million times faster. The speed of light is a million times faster than the chemical switching speed of our brains. So there is a mismatch. Ultimately we’re going to -- we’re in the process of reverse engineering our thinking. I have a new book called the Singularity is Near, when humans transcend biology, which will be coming out this fall. And I concentrate a lot on the process of reverse engineering the brain, which will be the source of understanding the software of human intelligence. Once we reverse engineer that, we’ll be able to take those methods -- we’ll be able to expand our AI toolkit to include the methods the human brain uses. But then we’ll be able to apply them to computer substrates that run a million times faster than biological thinking. When we have non-biological thinking processes working inside our own brains, they will work a lot faster.
FC: How would that work? The human brain is like a bottleneck. You’ve got a fundamental biological problem -- to remember something or make a connection that requires these biological processes is a hardware issue. Are we going to supercede our biological hardware?
Kurzweil: We will have non-biological processes. I don’t even understand now when I remember something where that memory comes from. I try to think of some actress’s name, and I’ll think of it or maybe I won’t be able to think of it, but if I do think of it, I’m not entirely sure. I’m not conscious of how I did that. So, one could imagine one’s memory working a lot better than it does, and a lot of it being non-biological.
FC: But to actually fundamentally change the biology of the brain?
Kurzweil: We get into profound issues of consciousness. I do think our sense of consciousness will start to encompass these non-biological processes. We already have a lot awareness of what we’re doing with our computers and we have a lot of lack of awareness of what goes on inside our own brains. There isn’t a perfect correlation of what goes on in our bodies and brains and what we’re aware of. So as we expand the processes that go on in our bodies and brains with non-biological ones that extend its capability, I think those will also be in the province of our conscious awareness.
FC: Mind blowing.
Kurzweil: Certainly mind expanding.
FC: You’re also working on a product to predict changes in the stock market.
Kurzweil: The company is called Fat Kat. Financial Accelerating Transactions from Kurzweil Adapted Technologies. I founded it about five years ago. We have a blue-ribbon group of investors. Vinod Khosla, one of the founders of Sun Microsystems. He and John Doerr run Kleiner Perkins. He is our lead investor. Mike Brown is another lead investor -- he’s on our board. He was chief financial officer of Microsoft for many years, and chairman of Nasdaq. And we have a number of other prominent high-tech people backing us. And the concept is applying my field, pattern recognition, to the stock market, particularly to short-term movements of stocks. And if you look at a particular stock it looks like an electric cardiogram. It constantly goes up and down. These little movements look random, and there’s certainly a very large random component to the movement, but its not entirely unpredictable because companies have relationships with one another. They own each other. They’re in supply chains with each other. They’re in industries with each other. There’s all kinds of influences. If you see certain movements in certain stocks, that’s putting out all kinds of ripple patterns and it ultimately will reflect itself in other movements in other securities. In fact the speed with which those reverberations or implications occur is speeding up with the more rapid dissemination of information. We look at data from 10-15 years ago, and we can see similar patterns today. But the patterns 15 years ago were moving much more slowly. Right now, there’s an announcement, and 30 minutes later it's old news because it's been all over the Internet. Whereas, 15 years ago, it would take days for the information to move around. So we actually see very similar patterns. They just move more quickly.
FC: That implies that at the core of this is information distribution.
Kurzweil: Certainly. Information affects people’s decision making and that affects people’s purchasing of securities. And we also see the effect of increasing interest in so-called quant investing. Quant stands for quantitative. The idea of investing using computers is called quant investing. So we can see some simple methods that worked five years ago that don’t work today because of the improved efficiency of the market. The arbitrage from the simpler methods has been rung out of the system. So we have our sophisticated pattern recognition model -- we don’t program it a priori with our preconceived ideas of how the market should work. It’s very much data driven, and it's building its models based on what it sees. But it has the ability to build sophisticated models of how financial data interacts with each other.
FC: I assume there are massive amounts of inputs.
Kurzweil: Anything that’s quantifiable. Certainly all of the tick data of market transactions, but also a lot of company fundamental data and economic data. Analyst opinions -- things we can quantify. It builds models, and the result is that it makes predictions. It’s constantly updating what securities will do in different time periods, ranging from hours to weeks. And the objective is not to be able to predict these things perfectly, but to predict them better than chance. And it turns out we can definitely predict these movements substantially better than chance. That puts us in the position of being the house in a casino. The odds are slightly in the casino’s favor. On any two or three rolls of the die, the casino may make or lose money, but over 50,000 bets, it reliably makes money because the odds are in its favor. Of course, the challenge for the casino business model is that the transactions are not free. It has to pay for the casino and the people that operate it. It’s what’s called trading friction; they have to make enough on each transaction to pay for the process. So we have the same issue. But the odds are in our favor because we can make predictions substantially better than chance. Our system places lots of bets. Each bet is fairly small relative to the size of the fund. Some transactions win; some lose. But more win than lose so the system makes money and it makes enough money to overcome trading friction. There’s different forms of trading friction: The actual transaction costs, the fees, slippage -- you go to buy a stock that’s supposed to be $50.30 and you end up paying $50.32 cents. That’s slippage. Recoil: if you try to make a really big transaction you’ll actually move the market. But our system works, we’ve been trading with real cash for 2.5 years. We make 80-90% annual gains. We plan to launch this year a hedge fund using our technique.
FC: What’s a very idealized idea of where this will be in five years?
Kurzweil: Our model is Renessaince. They make 45% a year and they manage $5 billion. They do it year after year. Last year they made $2.25 on $5 billion and they keep some of that as a fee and return the rest of those profits to the fund investors. Similar technology. That’s our model of success.
FC: Can it learn?
Kurzweil: Oh, it’s constantly learning.
FC: Your future will no doubt change or shock the system, society, business changes.
Kurzweil: These are gradual changes which are already underway. It’s not like nothing is going to happen and we’re suddenly going to wake up in 2025 to a different world. We get there a step at a time, and it's already started. We can already see the business models changing. It's not just one change. It's not just a case of a CEO who presides over a company that operates the same way year after year. Already companies need to reinvent themselves in order to succeed. There’s a shock when an industry resists changing its business model. The recording industry resisted changing its business model; they tried to keep the same business model that was around when my father was a kid. Selling an album with maybe only one or two songs that someone wants for a pretty expensive price. The bottom line is that industries have to change the structure of their business models. Very often it’s a new set of organizations that adopt a business model that’s consistent with disruptive change that displaces the old ones. But people aren’t necessarily going to keep the same jobs or careers for their whole life -- especially when we change the concept of the human life cycle.
FC: Do you think there will be a point where people can turn themselves on and off?
Kurzweil: Well, we will be able to separate the software of our lives from the hardware of our lives. That’s another advantage of non-biological intelligence. If you change computers, the viability of your software files isn’t lost. You can copy them over. They outlive the hardware. They don’t necessarily live forever though. If you walk away from some software files for a while and nobody cares about them and you come back, you may find it almost impossible to revive them. Try coming back now to some software that existed on some PDP-One eight-inch disk drive. You’d have to find all kinds of layers of hardware and software to revive that information. In fact, that’s a very serious issue with standards and software formats constantly changing. Software actually does require constant maintenance to remain alive. The basic message is software remains viable if somebody cares about it. And there’s an analogy in our own lives even today: if you don’t care about your own life, then you’re likely to not maintain your physical body very well. But yes, we’ll ultimately be able to separate the hardware and software of our lives. Right now, they’re deeply embedded with one another. When our hardware crashes, the software goes with it. And there is actually information in our brains. It is literally information. I’ve estimated it in thousands of trillions of bytes, reflecting our skills, our knowledge, our memories, our personality. You can argue about those estimates, but there is a certain amount of information there. And right now, when someone dies, that information is lost. In my view, death is a tragedy. It’s a tragic loss of all of that precious knowledge of experience and personality. And ultimately we’ll be able to separate the hardware from the software. But, as I pointed out, it doesn’t necessarily mean the software lives forever -- it’s just no longer dependent on one hardware substrate. It will only live as long as someone wants it to.
FC: What fields and industries will become less important in this future?
Kurzweil: We’ve already seen a migration away from jobs that involve extending our bodies. At the beginning of the 20th century, 30% of the population worked on farms and 30% worked in factories. Those figures are now down to 3% each. So we’ve seen a profound shift there already. Increasingly, professions involve expanding the reach of our minds and creating knowledge. Knowledge in very broad forms, whether the knowledge is music or art or culture or writing or science or technology. Increasingly that’s where our work efforts will be directed. I think people should go with their passion. If they really have a passion for art, we’ve seen a great empowering of the arts through technology. There’s a tremendous need for creating graphics and so on. I know artists that could hardly make a living who are now in tremendous demand as Web designers. It does pay to learn skills to be able to express ones passion in the vernacular and technology of the times. I do have exposure to a variety of fields, and it’s remarkable to me how technically sophisticated every field is becoming, from library science to music to art to certainly science and technology. I do think that we need to have more kids in America pursue science and technology careers. In Asia they seem to understand that. I have some graphs that show the number of science and engineering graduates in the U.S. is actually going down slightly. 60,000 10 years ago to about 55,000 today. Whereas in China, for example, it was only a fraction of our level 10 years ago, and they’re now up to about 300,000 engineers and scientists a year. We see similar progressions in India, Japan, and Korea. So those societies seem to understand that the cutting edge of future economic viability is science and technology and they’re preparing their kids for that. The counter-argument to that is that even our kids who are not becoming scientists and engineers are nonetheless actually becoming very sophisticated to technology. So you talk to a musician and he’s actually extremely knowledgeable about computers. That’s true with almost every field.We see one trend towards increasing specialization, whereas, take my field of pattern recognition: It's so diverse and there’s so many different areas it’s hard to keep up with even a small portion of it. On the other hand increasingly important work needs to be interdisciplinary -- to draw upon many different fields together. For example, the work I did in speech recognition, we had many different fields: linguists, speech scientists, signal processing engineers, mathematicians, complexity theorists, computer scientists. We had all these different fields working together so we need to be able to build bridges between these different disciplines.
FC: Is that an argument for specialization -- because you need to advance knowledge -- or an argument for the person or program that brings that knowledge together, assimilating it.
Kurzweil: Increasingly that’s in the entrepreneurial field, where to actually achieve something of value, you have to be able to combine different fields. Search engines had to marry library science with databases and intelligent search algorithms and so you’ve got mathematicians and linguists working together. Increasingly true of any important practical project.
FC: What would you like to be doing in 100 years?
Kurzweil: I do have a goal of being a successful 25-year-old female rock singer.
FC: How long have you had this goal?
Kurzweil: Awhile. I did a decent job at TED 2001. You can see the results.
FC: Was it a pre-existing goal?
Kurzweil: Yeah, I particularly like female singers. I realized it would be fun to be one. Of course, I always wanted to be with one, but to be one was also kind of cool. And I actually do hope to return to that goal. It’s got nothing to do with gender confusion. It has to do with the discovery that we do have different people inside us that we’d like to be. I really enjoy innovating; I get captivated with ideas. I remember at age 5 -- it wasn’t an idle fantasy, I was absolutely convinced I would be an inventor. As I grew a little older, 8, 9, I had all these different invention projects. I read the Tom Swift novels where the whole message was no matter what problem you got into there was some idea -- and you could find it -- that would overcome these seemingly overwhelming problems. And I continue to have lots of ideas. I do end up committing to some of them and when I commit to a project I really do see it through -- it may take awhile. But I’ve got a lot of other ideas I’d like to pursue. So I do see myself 10 year from now, 30 years from now, 100 years from now continuing to be immersed in the world of ideas and trying to make ideas real. It takes a lot of work and commitment and passion, but it is my passion. What’s exciting for an inventor is to actually see ideas come into the world and affect people’s lives. I call it the link between dry formulas on a blackboard and impacting people’s lives. That’s not the only way to be; a theoretical scientist is excited just by the idea. But what turns an inventor on is to actually have the idea get out into the world and impact people. So when people send me albums they wouldn’t have been able to create without the type of synthesis that we pioneered or I get letters from blind students who use our reading technology in education, that’s thrilling. It shows the power of ideas. So I see myself continuing in that path.
FC: You’re almost the anti-specialist. You’re envisioning an existence where you’re capable of integrating different fields.
Kurzweil: That’s true, but I’m also very committed to my main interest, mathematics and pattern recognition. I’m impressed with the power of mathematics to impact the world. You can have a set of mathematical formulas that can actually make predictions in the stock market or understand human speech or understand patterns in biological processes and overcome health problems. And at the core of that is mathematics and pattern recognition. I do think that most of our intelligence is based on pattern recognition. Human thinking is actually not very good at logical and analytical thinking. We are very good at recognizing patterns.
FC: Storage formats: Is there a solution?
Kurzweil: I’ve actually thought about that. I think it’s a fundamental philosophical issue. I don’t think there is a solution. Other than the insight that information will survive if we care about it. I have files that I have nursed along through different formats that are still alive because I care about them and I manage them. I also have some old files that I probably won’t be able to retrieve because I haven’t kept them up to date. To the extent that information will encompass more and more of our lives, ultimately our whole personality and intelligence can be seen and recorded as information this becomes a very important insight. Our whole survival will continue if we care about ourselves. I don’t see a technological solution. I’ve been looking for one because I actually want to create a database of all my files. I have hundreds of boxes of paper records. My father was similar; he has all of his letters and college papers and Ph.D. theses. I have 50 boxes of my father’s papers. And I want to scan it all in. But then, if I scan it all in and have this big database, what format can I possibly put this in that it would still be viable in 50 years from now? And there is no such format.