Google, Singularity University futurist Ray Kurzweil on the …

Posted: January 20, 2017 at 11:40 pm

Please Sign In and use this article's on page print button to print this article.

Sep 6, 2016, 4:18pm PDT

Leia Parker Managing Editor Silicon Valley Business Journal

Leia Parker Managing Editor Silicon Valley Business Journal

Ray Kurzweil sees a future in which we can connect our brains to the cloud to augment our more

Vicki Thompson

Ray Kurzweil is a futurist, a director of engineering at Google and a co-founder of the Singularity University think tank at NASA Ames Research Center in Mountain View. He is a nonfiction author and creator of several inventions.

Kurzweil met with the Silicon Valley Business Journal to discuss how technology's exponential progress is rapidly reshaping our future through seismic shifts in information technology and computing power, energy, nanotechnology, robotics, health and longevity.

Ray Kurzweil sees a future in which we can connect our brains to the cloud to augment our more

Vicki Thompson

This Q&A interview has been edited for length and clarity.

You've written about the potential to greatly improve humans abilities through a fusion of technology with biology. Could you describe how youre trying to motivate people to make this happen?

I've tried to articulate where I see the technology going and the underlying force behind it, which I call the law of accelerating returns, and the enormous opportunities created by exponential growth of information technology. It's not intuitive our intuition about the future is linear. But the reality of information technology is, it's exponential.

Exponentials are quite seductive because they start out sub-linear. We sequenced one ten-thousandth of the human genome in 1990 and two ten-thousandths in 1991. Halfway through the genome project, 7 years into it, we had sequenced 1 percent. People said, "This is a failure. Seven years, 1 percent. It's going to take 700 years, just like we said." Seven years later it was done, because 1 percent is only seven doublings from 100 percent and it had been doubling every year. We don't think in these exponential terms. And that exponential growth has continued since the end of the genome project. These technologies are now thousands of times more powerful than they were 13 years ago, when the genome project was completed.

Most importantly, we will be able to reprogram this outdated software that runs in our bodies, through biotechnology. We're now seeing clinical implications: It's now a trickle. It'll be a flood over the next decade. We're literally going to be able to reprogram biology away from disease and away from aging.

People say, You know, my cell phone is literally billions of times more powerful per dollar than the computer I used when I was an undergraduate, but it only applies to these gadgets we carry around. Thats not the case. Its going to transform food, printing, manufacturing of housing and energy.

Solar energy is growing exponentially because we're applying nanotechnology to the construction of solar panels and energy storage. It's now 2 percent of the world's energy, so people dismiss it as: It's 2 percent. It's a nice thing to do. It's a fringe player. That's not going to solve the problem. They are ignoring the exponential growth. Two percent is only six doublings from 100 percent. We're doubling every two years. That's 12 years. We can meet all of our energy needs through solar.

When I talk about radical life extension through biotechnology and nanotechnology, you can say, "Yeah, but we're going to run out of resources." But the same technologies that are going to extend life are also going to expand resources.

Ultimately, we can produce food extremely inexpensively through vertical agriculture, and we'll be able to print out everything we need through 3D printing. It's not ready yet for prime time. We're kind of in the hype phase now.

By 2020, we'll have sub-micron resolutions. We'll be able to print out and begin a really revolutionized manufacturing. We'll be able to print out modules and snap them together, Lego style, for construction of houses and office buildings. It's already started in Asia. It's not cost-effective yet, but these technologies have a 50 percent deflation rate.

We'll be able to print out clothing for pennies per pound with 3D printing in the 2020s. And there will be an open-source market of designs that are extremely inexpensive.

How is the rapid increase in computing power democratizing access and changing our economy?

I had saved up for years from my paper route as a teenager to buy the Encyclopedia Britannica for $1,000. I thought it was fantastic. It had all these incredible articles about everything I could imagine. Well, now a kid in Africa with her $30 smartphone can access a much better encyclopedia for free, and that's one of thousands of free fantastic information resources that are at her fingertips.

This is all factored out of the economic statistics. They say, Well, economic growth is limited. That's because we put this growth in both the numerator and the denominator. This kid in Africa that spent $30 on a smartphone is walking around with a trillion dollars of computation and communication and other intellectual resources, circa 1968, and still only accounts for $30 of economic activity.

People say, Okay, these fantastic comparisons apply to this strange world of great devices. You can't eat that, you can't live in it, you can't wear it. All of that's going to change with 3D printing, with virtual reality, with all these other resources that are expanding exponentially, and they sneak up on us. When these things start out, they don't work. By the time they work, they've been around for a long time, and they kind of sneak up on us.

What is your current focus in your work at Google?

I am a director of engineering, and I'm heading up a team working on natural language understanding. Language is like our most important method of communication. All of human knowledge is embedded in language. When we expanded our neocortex two million years ago and we got these big foreheads, the first thing we did was invent language so I could take an idea in my head, which is a hierarchical set of symbols, and transmit it to your neocortex. We needed a hierarchical medium to do that communication, so we invented language.

Since then, we've invented billions of documents in language with all of our knowledge. If we could actually understand the meaning of documents, that would unlock this great world of knowledge to computation and ultimately to humans so we can have our computer programs actually understand what they're reading. And we've already made great strides in that.

What's your primary objective for your work at Google?

We're part of an effort working with other teams to move towards an actual understanding of documents. So a search would not just be looking for keywords, it would actually look for meaning, and language translation would be based on meaning. It's a long-term effort to really understand language. Google's motto is, "We organize the world's information." Well, the most important information if you write a blog post is: What are you trying to say? You're not just trying to put together an interesting collection of words. Google's not the only company working on this, but that's a grand challenge to actually understand the meaning of documents.

What occupies your mind the most right now? Is it machine learning or another area of interest?

Well, I've been very focused on artificial intelligence for 50 years. I actually met with the founders of artificial intelligence. Marvin Minsky, who became my mentor, was the father of the symbolic school of artificial intelligence. And then in '62, when I was 14, I met with Frank Rosenblatt, who's the founder of the connectionist school and neural nets. He invented the first neural net called the Perceptron, and I've been immersed in that field for more than 50 years.

At the same time, I'm a writer and a futurist, so I keep track of all the world's technologies and how they're interacting.

I've had a long-term interest in health, which comes from, for example, my father dying prematurely of heart disease. That interest just comes from being a human being with a version 1.0 biological body. But that now has become an information technology, because we've unlocked the information basis of biology, which is genes, and have the meanings of actually reprogramming this outdated software. This interest, which was not related to my interest in computer science, has become now a field of computer science.

You serve on the board of Martine Rothblatt's company, United Therapeutics. What is that company doing in this area of health?

Yes, I've been on the board since that company was founded in 1999. That's one very good example of biotechnology. I've written about this for a long time, but now it's becoming a reality. We can actually print out hearts, lungs, kidneys, and populate them with stem cells and grow out a human organ. This is being done successfully in animals. We can do it in humans now with simple organs, like tracheas and windpipes.

We can do it experimentally with animals with more complex organs, like kidneys, lungs and hearts. That will be coming to a human near you in five to 10 years, but it's happening. If you can do it in a primate, we know we can do it in a human. We have to go through the whole regulatory and safety process to perfect the technology, but it's coming.

Youve also written about the importance of brain mapping. How does that factor into technologys exponential progress?

I track brain reverse engineering very carefully. We can do noninvasive brain scanning in humans. We can actually see now single inter-neuronal connections forming in real time and firing in real time. And there are a lot of different parameters that are important: the speed with which you can do it, the bandwidth and how deeply into the brain you can go with noninvasive scanning. But all of these parameters are rapidly improving.

How important is understanding how the brain functions in order to develop better artificial intelligence?

To me, the importance of brain reverse engineering is not that we're going to copy exactly how the brain works in cell rhythms, but find out its basic principles of operation. Then we can use good engineering to create the same principles, but do it more quickly with electronics. Our neurons transmit information using electric chemical signals that travel a few hundred feet per second. Electronics are already millions of times faster than our neurons, but we need to understand the principles of how it works.

In my last book, How to Create a Mind, I talk about the evidence we already have on how our neocortex works. It organizes 300 million modules, each of which can learn and understand a pattern, and they're organized in hierarchies. We create that hierarchy with our own thinking, and there have been a lot of insights from the brain reverse engineering projects that really support this thesis.

You've given timelines for bringing on a transhuman reality, in which our capabilities are dramatically increased through the power of technology. How are we doing in keeping to those?

We're very much on schedule. Artificial intelligence itself has done remarkable things that people didn't expect to see for a long time, like drive cars, like play Go better than any human and understand language to some extent.

Jeopardy is a language game. Watson got a better score than the best two humans combined, and answered this query correct: A long, tiresome speech delivered by a frothy pie topping. It quickly said, "What is a meringue harangue?" That's pretty good. And Watson got its knowledge by reading Wikipedia and other encyclopedias. It doesn't read as well as you or I, but it reads a lot more documents. It read 200 million documents. We can't do that. It was able to combine all of its knowledge from that effort.

We're making tremendous progress on understanding the brain. I think we're very much on track to have human-level AI by 2029, which has been my consistent prediction for 20 years, and then to be able to send nanobots into the brain in the 2030s and connect our biological neocortex to synthetic neocortex in the cloud.

This is impressive by itself, but it's more impressive because it connects to the cloud. If you do speech recognition or intelligent search, it goes out to the cloud and makes itself a million times smarter. It does that without you even being aware of it. People don't even know it's happening.

We can't do that directly from our brains yet. We do it indirectly with our devices. We have to use our fingers and our eyes and so forth. Ultimately, we'll do it directly from our brain and not just do search and translation directly from our brain, but actually access synthetic neocortex. So just the way this [he holds up his smartphone] makes itself smarter by connecting to the cloud, we'll make ourselves smarter. And that's the ultimate application of artificial intelligence: to extend our mental reach. That's a 2030s scenario.

Your Singularity University co-founder Peter Diamandis has told me he believes that today, its possible for people to live long enough to live forever because of these rapid technological changes. Do you anticipate this could happen for you?

I'm planning on it. So far so good.

You're 68 years old now?

Yeah. And I could be hit by the proverbial bus tomorrow, but we're working on that, too, with self-driving cars.

What would it take to dramatically extend the lifespan of humans?

I think we're on the order of a dozen years away from a tipping point where we're adding more time through scientific progress than is going by. People say, You think you're going to live hundreds of years taking these supplements, and with your lifestyle and so on, that you describe in your book? And I say, No, the goal of that, which we call Bridge One, is just to get to Bridge Two, which is the biotechnology revolution. And a dozen years from now, we will really have arrested most disease and aging processes. Not all, but we'll reach a tipping point where we're adding more time than is going by.

And then Bridge Two will be a bridge to the nanotechnology revolution: medical nanorobots that can augment our immune system and go beyond our immune system. Our immune system evolved when it was not in the interest of the human species for us to live very long, so it did not select for long life. It doesn't work on cancer for example. So we can finish the job with medical nanorobots that can basically defeat all disease and aging processes. That's 20 years away.

How would that help to bring about a period of abundance?

Well, that will enable us to live longer. Then people say, "We're going to run out of resources." That's where abundance comes in. Solar energy is doubling every two years because we're applying nanotechnology. We're only six doublings from meeting all of our energy needs through solar. We have 10,000 times more sunlight than we need to do that with. We'll have 3D printing for modules to snap together and create a house, for food, for clothing. We'll meet our physical needs through 3D printing. We'll have virtual realities, so we won't have to travel as much. So ultimately, we will have an age of abundance we won't run out of resources.

What would people do with themselves?

We'll continue to create knowledge. What do we do now? Sixty-five percent of all jobs in the United States, Europe and Asia are information jobs. It didn't exist 25 years ago. So what if people are creating art for websites or creating music?

We have 15 million college students and 15 million people that service them. That's 30 million people. It was 65,000 college students in 1870, so we're moving up Maslow's hierarchy. We're doing more gratifying things: creating knowledge of beauty, like music and art, science, technology.

Are you worried about individuals' worst impulses potentially throwing a wrench into the works?

Well I think we're getting better because I think communication has democratized the world. You could count the number of democracies in the world on the fingers of one or two hands a century ago. You could count the numbers of democracies in the world two centuries ago on the fingers of one finger.

We certainly don't live in a perfect world, but this is the most peaceful time in human history. People say, "What are you kidding? Don't you pay attention to the news? Didn't you hear about the incident yesterday and a week ago?" Well that's the point. Our information about violence and what's wrong with the world is getting exponentially better. It could be a battle that wiped out a nearby village and you wouldn't even hear about it a century ago. Now, there's an incident and we not only hear about it, we're immersed in it, we experience it. That's painful, but it's actually a good thing because it motivates us to do something about it.

Steven Pinker's The Better Angels of Our Nature documents an exponential decline in violence. We rightfully get upset about incidents that kill tens, hundreds, thousands of people. You don't have to go back that far in history to see incidents that killed millions tens of millions of people. It's not like this type of violence and intolerance didn't exist. We just didn't actually have very good information about it a century or two ago.

Currently, we're in the political season, and weve seen plenty of polarization. Where do you stand with respect to the U.S. presidential election and how it has developed?

Technology is a double-edged sword, and it can also spread intolerance. I'm not happy with the level of intolerance that we see expressed in some parts of the political sphere. But I do think without commenting specifically on the current presidential race there's a world consensus on tolerance, equality, democracy, liberty, and then we complain about the extent we see things that don't live up to that. We're moving in the right direction. History is always a messy process, and we have much better information about the mess now than we ever did before.

Join the conversation: Follow @SVbizjournal on Twitter, "Like" us on Facebook and sign up for our free email newsletters.

Originally posted here:
Google, Singularity University futurist Ray Kurzweil on the ...

Related Posts