Markoff: Morning. So I’m going to introduce our panelists and then I’m going to introduce our topic and we’ll get going. Philip Zelikow, who’s at the end, is the visiting managing director at the Markle Foundation. Steve Jurvetson, who’s sitting next to me, is a partner at Draper Fisher and Jurvetson. Jessica Rosenworcel is a commissioner at the FCC, and Ken Washington is VP of research and advanced engineering at Ford Motor Company.
So for a reporter, it’s kind of the best of worlds when you have one set of facts and you have 20 different opinions about what those facts mean. And when it comes to technology and jobs in America, I think that’s kind of where we are, and I just wanted to frame things very briefly before we get going. So we’re in the midst of our second meaningful debate on the future of technology and jobs. The last time the nation really went at it was briefly before the Vietnam War, in the 1960s, and that really grew out of a conversation that Norbert Wiener, who coined the term cybernetics, had with Walter Reuther in the early 1950s, and there was this long delay because of the Eisenhower administration and then the Kennedy administration. We started to talk about technology and jobs, and then the Vietnam War came along and it kind of went away. And now we’re back, and we’re back with a vengeance, and it kind of feels like “Rashoman” or “Blind Men and the Elephant.” You know, you have a world in which you have a Rice computer scientist like Moshe Vardi who will argue that by 2045 there will be no jobs—there will be no jobs, computers will be doing everything, and he argues this with great certainty. And on the other side of the equation, you have the International Federation of Robotics that’s arguing that robots are going to bring the greatest job renaissance in history. Now, you have to sort of square that circle, and I think that people like David Autor have gotten a great deal of attention for arguing that the middle is dropping out of the economy, sort of that the plug and play nature of the Internet has taken out the class of our society and it’s largely gone, and that the job growth is at the bottom and the top of the economy. And then at the same time, there’s recent NBER research that shows that the top’s kind of dropping out too, and there’s this kind of downward slide—it’s not a trough, but it’s the slide downward as skilled workers, who are being pushed down the ladder, take the jobs of less skilled workers. And, you know, just sort of a cautionary note before we get going, I wanted to say, I have a friend, Paul Saffo, who likes to say never confuse a clear view for a short distance, and I’ve been around Silicon Valley long enough to believe that there are a lot of visionaries that say things and some of them happen and some of them don’t. For example, Jeremy Rifkin wrote this book called “The End of Work” in 1995, and in the ensuing decade the economy added 22 million jobs. So, along those lines, there are two recent books that have both drawn this relationship between the 13 employees at Instagram and the 140,000 employees at Kodak, and they’ve made the linkage that in some way the disappearance of those 140,000 jobs was related to the 13 jobs at Instagram, and we can all see in just a second that it can’t be that simple. The Instagram jobs did not emerge until there was a mature Internet, and so how many jobs did the Internet create before it allowed something like Instagram to emerge? I think that McKinsey estimated in 2011 there were 1.2 million jobs created by the Internet before there was an Instagram that could come into existence. And at the same time—also you know, if you know anything about Kodak, it probably wasn’t the Internet that killed Kodak. They contributed a lot to their own demise. Fuji, for example, did just fine thank you navigating this kind of technology change.
So let’s explore the impact of computing and jobs over the next decade, and let me start by asking Steve, because I think we might have a good discussion here on the question of jobs and technology, where do you stand? You know, between the Rice computer scientist and the International Federation of Robotics, where are you?
Jurvetson: Well, I’m sitting right here. And I believe that we are in an era of profound change, and we may be at peak jobs, we may be heading to an era of abundance where people don’t need to work and robots are like our serfs and indentured slaves of past, allowing for enlightenment and science and this—and there’s a nirvana one could imagine if you think far enough in the future. Perhaps the counterpoint of Saffo’s is don’t mistake a clear view for looking far enough in the future, and I think if you look out 500 years, it is inevitable that robots could do any physical task better than a human that is repetitive, that, you know, art, and the one-offs and the organic farming for fun, that will continue forever, but the bulk of the economy will be robotic. Autonomous cars will take over for the Uber drivers, the places in any manufacturing line that still have humans doing a repetitive task will disappear, without question. Just look far enough in the future, I don’t think it’s debatable. What is debatable is that five years, 10 years does it matter in the near term? And that’s a different question.
But if I think of the long arc inching back, I see a future where there’s profound change and every business becomes an information business. So whether or not there are jobs, those jobs will be information work, they won’t be manual labor. If that’s the case, it’s kind of related to things Jaron was just telling us about in the prior talk. We will all be competing globally. Regional shelters of competition will disappear and the basis of competition will be something that looks like software or entertainment. Not everyone will want to do that, and the people who do have a vastly different skill base. The top decile of programmers, even at Google, are much more productive than the bottom 90 percent of programmers at Google, and you’ll have a further acceleration of a power law in income distribution and a technology accelerated rich-poor gap in any case.
Markoff: So why hasn’t the acceleration shown up in productivity numbers yet?
Jurvetson: Which part, the acceleration of technology?
Jurvetson: Well, so productivity—so you could look at various economic metrics and then you have to double click on what are we talking about. So for example, Peter Thiel talks about median wages in America being stagnant since 1975, but not the average, right? So in other words, the average wealth has gone up, but it’s just concentrated in the rich. So if you say productivity, well, what kind of productivity? We are productive as a species. Everything that the Internet has brought, to your earlier question, makes us more capable than before. We all know that, we feel that, and then some economist give you a statistic that seems completely dissonant with reality, right? So I think we are more productive.
Markoff: Philip, I wanted to—so let’s assume for a second that Moshe Vardi is right and there is this dramatic impact on jobs and the economy. How do you manage from a—how do you respond from a policy point of view—not from a market point of view, but what can we do beyond what the markets are already doing to try to cope with these changes?
Zelikow: First of all, we should not accept the premise that the jobs are going away and it’s the job of government to manufacture Band-Aids and have partisan fights over the scraps. The job of government, and frankly the job of the private sector is to realize that we’re in the early stages of a digital revolution that is on a scale comparable of the Industrial Revolution of more than a hundred years ago. What happened more than a hundred years ago? America remade itself comprehensively on a broad front, on a dozen different issues, from universal high school to the electrification of industry and a dozen other agenda items that transformed the country so it coped magnificently in the industrial age of the 20th century, and today we have the legacy institutions that thrived in the Industrial Revolution and were highly functional for that revolution increasingly dysfunctional in coping with the early stages of a transformational change. What does that then tell me? It says we need to adapt again. We need to have the same kind of broad agenda that we had before. Why am I not giving up on employment? Let me give you three examples. One, imagine the decentralization of production, industrial production, inherently centralized, hierarchal, standardized for mass production, networked production inherently decentralized, customized, right? So imagine first the notion that the factory system becomes not hundreds of micro factories but potentially thousands of micro factories, even mobile factories, all of which are employing people and moving production closer to the market. The supply chain’s moving closer to the market.
Example number two: Think about how you empower the frontline worker at the edge of the network. The home healthcare aid who can seem to have superpowers if they’re wearing a DAQRI helmet or Google Glass and are connected to a medical network and are using health sensors, and thus have the kind of technology that we now give to 19-year-old frontline soldiers in Iraq, which could also go to 23-year-old home healthcare aids who now make $10 an hour and are in one of the fastest growing professions. By the way, that happened before, in the Industrial Age, where we converted factories from steam engines to electric engines, from group drive to unit drive, and common laborers, which was much of the workforce, became people like machinists, and businesses lobbied to have universal high school so that they could all read and write. That’s the kind of agenda then that comes out of that different vision of how you empower the frontline workers at the edge of the network, in retail, in healthcare.
Example three: Think now about what we imagine as the non-tradable part of the U.S. economy, the 63 percent that’s in services. Imagine, however, that a fraction of that becomes tradable through digital commerce and telepresence, so that a patient in New Delhi walks into a doctor’s office in Richmond, Virginia to be diagnosed about a mole on their wrist using sensory technology and remote technology, and this is not just employment for doctors, of course. It’s employment for medical assistants, nurses, and all the rest, and a vast industry underneath that.
Now, I’ve just cited off three examples where, if we have the vision to think about it, none of these are science fiction. People in this room know the technologies that I am talking about are by no means in the distance, and none of this is crippled by a shortage of money. America is glutted with private capital, and money has never been cheaper to borrow. What’s needed then is a broad vision of America’s economic future that is not dismal and handwringing and worrying about the Band-Aids for all the people who will be left behind, but is exciting on a dozen different fronts about how do we rebuild the American dream for the digital revolution.
Markoff: So I’ll come back to—I want to poke on that a bit, but let me go first to communications and then transportation. So a couple of years ago, there was national academy—this is a question about spectrum and how does spectrum play into this. And I would have thought maybe not, but then there was this national academy study that suggested—so right now, we auction off the spectrum and we give it to one buyer who can use it for a long period of time, but there’s this technology that in principal would make it possible to auction spectrum on a microsecond basis. We could use the spectrum in radically different ways. And the national academy report was so optimistic, they suggested it would be so productive that it would pay off the national debt. That caught my attention. It would also, they argued, sort of create the environment for another Internet sort of ecology in which you could create new ideas and innovations and technologies. So is the FCC creeping in that direction? You’ve talked about it—is that 5G?
Rosenworcel: Well, I think the most important zoning that goes on in this country right now is not on the ground. You can’t see it. It’s in the airwaves above, because nobody in this room has left their home this morning without at least one mobile device, and if you did you’re probably cursing that you left it somewhere along the way, maybe even strategizing about how you return it—get it back. But what’s really important is that the FCC finds a way to zone our airwaves so that all of that wireless activity can occur at once. And just like on the ground you wouldn’t want to have an industrial use next to a residential neighborhood, up in the skies we need to make sure that broadcasting and broadband can coexist, and the more that we demand of our devices and the more that we speed into the Internet of things and put sensors on everything, how we zone those airwaves and how we optimize this mobile universe is really important. That is going to take a mix of things: traditional licensing, like you mentioned, where we just sell off the spectrum to the highest bidder and ask the private sector to put it to use; it’ll take unlicensed spectrum, like Wi-Fi, where we create a public commons where people can experiment and get online; and it’s also going to have to take new and creative uses of our spectrum that’s more dynamic, because as we move from 4G to 5G technologies, we’re going to be able to identify use cases that are only occasional and we can find ways for that spectrum to be shared.
Markoff: Is 5G more than a number now?
Rosenworcel: Well, let’s see, 4G spectrum today—the U.S. is doing a terrific job. We’ve got about 5 percent of the population but about half of the world’s 4G LTE deployment. The worst thing that we could do is rest on our laurels. The rest of the world is starting to try to identify what the next generation of wireless service looks like, and that’s called 5G. We’ve got activities in China, South Korea, and the E.U. right now actively trying to identify what the next 5G standard looks like. So I think it’s time for the U.S. to get into the game, and I think the way we’re going to do it—and here’s where I’ll get spectrum nerd on you—is instead of just looking at spectrum between just 600 megahertz and three gigahertz, which are the traditional bands for mobile spectrum, we’re going to have to look way, way up to infinity and beyond. We’re going to have to look up so far it feels like the stuff of a cartoon movie.
Jurvetson: But then it gets really narrow. It’ll be like a laser.
Rosenworcel: Then what we’re going to have to do, is we’re going to have to really widen channels, and we’re going to combine them with lots of small cells, which are micro-towers, not those big towers we complain about in our backyards, but lots of little micro-towers. And what that will do will give us in small, dense areas, particularly cities, incredible amounts of throughput for mobile activity.
Markoff: Is that spectrum licensed yet—I mean not licensed, is it under any kind of regime? Is it out there?
Rosenworcel: Not much. So think of it as fallow land, and it’s got lots of opportunities. It’s very exciting.
Markoff: Yes. So Ken is recent to Ford Motor company, but he’s not recent to autonomous vehicles, it’s just they didn’t go on the ground. They went underwater. He was at Lockheed Martin before, so he’s an expert on autonomous vehicles, but not the ones that are involved in carrying people around. So let me ask about the tradeoffs. So Sebastian Thrun, who was a pioneer in the Google Car, talks about saving 30,000 lives a year, but if we actually had his self-driving cars, we might un-employ three million people who made their living from driving cars. How do you see the future of cars in terms of whether they drive themselves or not?
Washington: So this is an exciting topic for us and everything that’s been said so far plays very neatly into my view, and Ford’s view of autonomous vehicles and the future of things regarding mobility. Because if you think about it, this is a pretty exciting time to be alive. We’ve got technologies all maturing, coming together at a time and a place, happening at a pace that’s never—that’s unprecedented, quite frankly. You’ve got the Internet of things coming alive. Sensors are cheaper than ever. The autonomous capabilities are no longer just things of science fiction and research projects. And so we’re very excited about taking these capabilities and knitting them together to build cars that have the ability to assist drivers to be better drivers. And I think this notion that one day you’re going to wake up and go to the dealer and buy an autonomous car is flawed. I think that—in fact, you can buy a car today that has many technologies that are already in it that are quite amazing, that will assist you with staying in your lane, that will alert you when someone’s in your blind spot, that will help you park the car and see around corners, and of course in the rearview mirror view and so on and so forth. Those will continue to improve over time as we continue to make investments in them. And I think this notion that it’s going to displace drivers or displace jobs is also flawed, because the technology has to be developed by someone, and we’re finding that we’re bringing lots of knowledge workers to Ford, and I think our competitors are doing the same thing, so we’re employing lots of people to develop these technologies and to mature them and to conduct research programs that will lead to a future where these autonomous driving capabilities are even better, and in certain controlled environments, in well-controlled environments, you can have very capable maps that map in three dimensions the roadways and the objects on those roadways. We’re convinced that there’s a future where cars can operate autonomously, but I don’t think that correlates at all to a future where you’re displacing people that have jobs.
Markoff: And what about the generational question? I mean, you’re new to Ford, so it might not be fair to ask, but there’s sort of a common wisdom thing happening that we have a generation of kids who don’t want to own their own cars. How does that affect Ford?
Washington: So we’ve noticed—we’re recognizing that the notion of mobility is getting redefined, and I don’t think that’s just young kids. I think the move to urban environments and the population boom that’s happening in megacities is leading us to think about mobility differently. I mean 20 years ago, you thought about vehicles in the context of one day I want to get my driver’s license, buy a car, and then be able to operate it. Now there’s a full spectrum of mobility modes that people think about, regarding I need to be able to move from point A to point B in this context, and that might not happen in a car. That might happen on public transit, that might happen with an Uber kind of service or a car sharing service. That might happen with a combination of riding a bike and a car. So we’re doing a number of mobility experiments to help us understand these modalities of mobility so that we can be prepared to deploy solutions so that customers will have an experience to move from point A to point B regardless of whether they have a car or not.
Now, we’re convinced that a car, an automobile will be part of a broader system of mobility solutions which are assisted by autonomous capabilities, vehicle to vehicle, vehicle to infrastructure connectivity—that’s where spectrum comes in. As I said, these technologies are all weaving together into an inflection point where once the standards are put into place and the policies are set and the technologies mature, you’re going to see a future where cars can communicate with each other, they can communicate with the infrastructure, and they’ll have built in and beamed in connectivity for not only assisting a driver to have a much different, much richer experience, but a safer experience, and also a totally different set of options for how you’re mobile in an environment.
Markoff: So if the self-driving car is in the nonspecific future—we now have a suite of technologies in cars that you just mentioned. What comes next? I know GM has talked about super cruise, cars that drive themselves at freeway speeds. Is that the next sort of, for standard consumers, things they can buy in their cars, what’s the next sort of assisted or augmented technology?
Washington: I think what’s next is, you’re going to see—every car company has their own branded version, and for Ford, we’re focused on additional driver assist technologies, and you’re going to see in our vehicles a richer set of driver assist technologies like adaptive steering, like adaptive cruise control, like additional cameras and sensors on the vehicle so that you have richer and a more reliable set of assistance and warning systems. And parallel to that, we have a very active research program to look at the next generation of autonomous vehicles, assisted by the next generation of sensors, including LiDAR sensors, with a very rich backbone of data analytics. Because once you have sensors that can map the world around you in three dimensions, the next challenge is, well, how do you process the data so that you can make the decisions, because the decisions that need to be made involve information coming from a myriad of sources, and that’s why it has to be fed by a very rich set of analytical capabilities, and we’re working on that as a very active research program.
Markoff: So I don’t want to give up on the jobs and technologies just yet—
Jurvetson: I was wondering if I could argue against it, specifically. Well, I think we just heard a number of examples that, if I understand Philip and Ken, they’re trying to argue that this won’t destroy jobs, or at least radically shift who got the jobs. And I think all the examples mentioned are perfect examples of the profound change that we’re having. So take the first one, 3D printing and the democratization of manufacturing, whether it’s print at home, whether it’s a digital biological converter that allows you to make drugs and biologics from code, which companies are pursuing. That’s great, but that’s a perfect example of taking a physical activity and moving it to the machine, right? The average person isn’t building a 3D printer, the average person is printing something. And in so doing, matter becomes code, and everything will cost a dollar a pound—
Markoff: You’re talking about displacing labor?
Jurvetson: Right. Because if everything costs a dollar a pound, all you’re paying for is the design, and then you have all these distributed printers to just print it, so it moves it to an information economy. Or take the example of telepresence of service economy. Doctors will go away, 80 to 90 percent of doctors will be replaced by an expert system, because the telepresence will be through a phone. And then the cars will destroy the Uber drivers. I think it’s a failure of vision to say they won’t.
Markoff: Okay, but wait, you’re still not disagreeing with Keynes. Remember, Keynes said that technology destroys jobs, not work.
Jurvetson: Absolutely. Meaningful work will flourish. It won’t be paid. And all of Uber drivers—I’ll get to that in a sec—but the Uber drivers aren’t going to become roboticists, right? Maybe 1 percent will. They’re not going to flip from that part of the economy to “I’m going to program my own autonomous car.”
Markoff: Phil, could you respond to that notion?
Zelikow: Yeah, because it’s actually—it’s basically applying technology to today’s models of how we serve people. But the models of how we serve people, like everything else in the networked world, are becoming infinitely individualized. So if I’m basically—what you’re seeing, in retail for example, is the proliferation of stock keeping units, of SKUs, to the point where they can’t keep products in inventory anymore because the demands are getting so fragmented. This is one of the things that’s moving supply chains to the market. So the notion, well, the labor simply consists of writing the code—what about the interactions with the humans? Now, if I’m coming in, I want to buy a shoe that’s personalized to me, you know, and I want that shoe actually made for me, maybe even on site, the notion that that’s all going to be done in an interaction with machines I think just assumes a model that is going to be an inferior model of customer service. We’re talking about models of customer service, the creation of new artisans and new kinds of makers at a completely different level than anything we can imagine right now. So rather than impose basically technological innovation on today’s model of centralized, standardized, and relatively impersonal customer service, instead, imagine that everyone can be served when they want, how they want, with human assistance where they need it, which I think they will need, and now you’re talking about a whole different range of empowered frontline employees. And what we saw in the adaptation to the Industrial Revolution is people said, by the way, when they were introducing electrified industry, they said, “These laborers can’t handle having electric drills at their own workbench, they’re too stupid. The workers aren’t smart enough.” And so what happened then is that the employers then lobbied for high schools to be built that would train the workers better, so you actually had a demand for a more educated workforce and then a supply of more educated workers, which transformed the whole meaning of blue-collar work in America. A hundred years ago, 30 percent of the workforce in America identified themselves—identified themselves—assimply laborers, and we were employing millions of children. That was the workforce of 100 years ago. Then we created basically a completely different workforce, that, by the way, then had to be trained and educated in completely different ways. Think about the opportunities to train and educate the workforce that are now emerging for us and the kind of jobs and empowerment that they could have, and I actually think it’s premature to assume that that future for labor and service is not bright in that more artisanal world.
Markoff: So one more question before we bring the audience in. Everybody seems to be saying that there is this vast new inequality emerging from Western society, from Piketty on down. What role does technology play? Is it all negative, or can it be positive in terms of fostering or—
Washington: Well, I’ll jump in. I have every reason to be encouraged by the role of technology, and I’ll give you a good example. As part of my onboarding process, I spent some time, for the first time ever, in an automotive manufacturing plant, and it was eye opening for me because you saw, in that plant, some of the most amazing technology. But you also saw—I’ll call them knowledge workers—knowledge workers in the plant, interfacing and interacting with the technology, manufacturing state of the art vehicles, and they were doing the part of the job that machines are not very good at. They were adding quality checks, they were doing complex integration moves that machines are not very efficient at, and they were proud of the work that they were doing, because they were manufacturing a car that they believed in. And I think there’s an analogy to that in the broader community, because as machines and robotics and automation become more ubiquitous in our society, I think you’re going to see the American workforce step up to the challenge of how do they interact and interface and educate themselves to add value in that ecosystem.
Zelikow: John, if could interject, there is a dichotomy here that just has to be addressed, and the dichotomy is, it’s man versus machine. What history taught us, in fact, is the complementarity of humans with the machines, humans empowered by the machines that then re-envision the society around them in new ways that are actually difficult to foresee. So instead of basically, for instance, in education where I work, seeing this as, “This is a battle between personal teaching and online teaching.” Wrong. False dichotomy. That’s the first generation dichotomy. The next generation dichotomy will be blended learning that actually empowers human beings, caring human beings in one on one connections, who now have access to a world of available knowledge, integrated into new designs for education. That’s what’s possible. Now, I’m not a techno-utopian, I’m not arguing that this is all bound to happen and we can just step back and do nothing. I’m saying that it’s way—it’s premature to become resigned either way. It’s actually, this is a society we can build and make, inspired by the example of history where we’ve done this before.
Markoff: To Ken’s point, Toyota also has brought people onto the line, because they found that when they pushed people out of the production process, the production process stopped improving, and the only way they could rethink production—so it’s going to be a while before there are machines. I mean now we’re in your 500-year timeframe. We’ll take that role and improve production processes.
Jurvetson: And by the way, I am a techno-optimist. That may not come across on this particular panel, but what I do all day is—but the question was does technology affect the rich-poor gap. I think it dramatically accelerates it, because all of the industries in the developing world that look like information businesses have a huge power along income distributions. The U.S. is one of the biggest wealth disparities, and it’s growing and China is rapidly catching up. It’s I think an inevitable effect when you have preferential attachment in a network economy that the big get bigger and you get a power law distribution. So both between companies—how many Google competitors thrive? Almost none. And then within companies—where’s the income within, at Google? It’s sort of an obvious, I guess, megatrend. So then the question is, is that bad? And that then relates to issues like what positive feedback loops there are. All of the issues of is it a bell curve or a power law, and then the only positive I can see is that all of the rapid billionaires that I’ve come to know in Silicon Valley, who’ve gone from like zero to billions in less than five years, tend not to ascribe their success to hard work, or even skill as much, because it rings hollow and they realize the an element of luck and their place in the world plays a large role, and they tend to be more philanthropic than just about anyone else you’ll find, and they give almost all back through—that’s the only relief valve I see to an otherwise ever-polarizing phenomenon.
Markoff: We just have a couple minutes. If there are questions or comments from—yes, in front.
Audience: What I think I’m hearing across is a couple of issues coming up. One is what is the role of humans in the future of technology, and is creativity—what do the robots make, or what does the technology make, and judgment, maybe the two characteristics that we should be educating for, and how does that affect our education system. And then I think, Steve, to your question, I ask how do you look at distribution of wealth, and if people are going to work less, you seem to have some theories about that. So those are the two pieces.
Markoff: Do you want to take a shot?
Jurvetson: Well, quickly, I think the distribution of wealth, in democracy, in the capitalist system, is going to polarize, and that is the inevitable long-term trend. And then the question is, is that horrible or do you have a social norm where, if you’ve got more GPD than most nations, you do something other than party like a rock star, and I think that luckily is the norm that’s sweeping the—you know, Bill Gates Foundation and others, who do amazing work, put out $4.4 billion this year in grants. So the notion of how we can affect the trajectory, I’m not sure. It might become a policy issue, it might become something radical and new. What I’ve tried to do proactively as an investor is think about the bottom of Maslow’s hierarchy of needs. The food, shelter, and clothing I think are being handled, online free education I think is being handled. Free medical care for everyone forever is not, so that’s something I want to invest in, to try to address what might be a really interesting transition phase. But I think the vector is just inevitable.
Rosenworcel: Well, you know, I think I want to touch on something Philip mentioned earlier. One of the great legacies of this country in the 20th century was the expansion of education to everyone, but if you look at where we are right now, that educational system is still built on this industrial era norm. If you look at the jobs that are available in the new economy, the gap between those jobs and the ways that we’re teaching right now is impossibly large. And one of the things that I work on extensively at the FCC is a program called E-Rate, and it’s got the goal of getting a gigabit to all of our schools in the not too distant future, because why should my kids learn today off of a textbook? That’s a lack of imagination. We should have interactive activities at all of our classrooms, at every level in this country, and we should not make people more unequal because of the technology but develop the skillsets that will make them complement those new technologies and participate better in this new economy and benefit more broadly from those new technologies.
Markoff: Thank you. One last comment.
Zelikow: Quickly, just for folks to get a sense of optimism, the Amazon of the late 19th century was called Montgomery Ward. Retail models evolve. And, by the way, Montgomery Ward was bitterly hated by all the merchants who thought it would drive all local merchants out of business. So these things are dynamic. We can shape them.
Washington: One last word on the question of innovation. I think it’s important for us all to remember that innovation is ultimately a human endeavor, and so no automation or robotics or machines will replace the human value of the kind of innovative collisions that occur when you put people together that have bright ideas, and that’s what gives me hope of the future.
Markoff: A great point to end on. Thank our panel, please.