Session Description: Relentless automation and the shift to platforms is changing how we organize and categorize work. As the economy evolves from “jobs” to “tasks,” policies, safety nets, and systems struggle to keep up. How do we ensure everyone benefits from the new landscape of work?

The full transcript is also available as a PDF, here.

Petre: We’re going to try to get you leaning forward rather than back and snoozing. And I’m going to ask my panelists to just come out and then I’m going to do a little bit of a setup.

The nature and future of work has been an ongoing topic of discussion at Techonomy over the years. So I want to throw out 10 quick facts to focus us on the moment and the issues that we face right now about it.

Number one is 10 years ago was the start of the Great Recession. In fact, it has its birthday next month in case anyone wants to plan a party. But since that time the economy has sort of clawed its way back toward prosperity and in the process, 16 million new jobs have been created in a workforce of about 160 million people. That’s number one.

Number two is we’ve gone from peak unemployment of about 10 percent in this cycle to what the economists euphemistically call full employment, which means only 4 percent of people are unemployed.

Number three, even more good news, is that after many years of stagnation, wages are up. Who knew? Median wages in full-time jobs have actually trended higher for the past couple of years. And yet, as we’re all aware, there are lots of crosscurrents and instabilities in the workplace and in the workforce, which is why we’ve titled this panel the way we’ve titled it, “The Uncertain, Unstable, Changing Nature of Work.”

Number four is that most, maybe all, of the net job growth is in so-called contingent jobs. It’s basically contract workers, independent freelancers, on-call workers, help agency workers. And by the way, did you know that more and more American families are moonlighting as well?

Next fact: The New York Times ran a piece a few weeks ago that 1.5 million workers have disappeared from the workforce. Presumably they dropped out of the workforce with the recession and just haven’t come back, but nobody has quite pinned it down yet.

Number six is despite the median wage increase, there’s something to what the Atlantic Monthly recently described as “slow, continual downward pressure on the value and availability of work in America.” That’s mainly as expressed in full-time jobs.

Next point: employers—including, I’m sure, companies represented here in this room—have been complaining and struggling to find workers that have the right skills for the jobs they need.

Number eight: there’s this ongoing and very mysterious sluggishness in new business formation and job creation. And I hope we come back to that; I think it’s a central topic. And I know people here in our group have a lot to say about it.

Number nine, there’s a phenomenon of what might be called ‘taskification’ going on. All up and down the work spectrum, from professionals to white-collar, to blue-collar, to pink-collar, to no-collar jobs, there’s fragmentation of jobs into tasks, and some of those tasks are eminently automatable and done better by machines.

And then finally, there is this issue of the morale of working people, also all up and down the spectrum. In a recent report into it, [one] quote: “A high baseline level of insecurity surrounding jobs, income, and benefits among American workers.”

So if that’s not a confusing picture, I don’t know what is, but that’s why we’re here today. Some of these effects are cyclical. Some of them point to permanent shifts in the nature of work; many of them driven by technology. But at this moment of boom and of full employment, it’s really hard to tell things apart, and that’s why we’re going to have this panel—to try to tease apart some of the strands, to surface some ideas, and see if we can shed some light.

Each of my panelists is very accomplished in his or her own right, plus each person on this stage has deep expertise in the question of work from a completely different facet from anybody else on the panel. So we have five completely different outlooks. And we’re going to see those compare and contrast as we go along.

So let me just quickly introduce people and then we’ll get going. So let me start with Paul Roehrig, who is the Chief Strategy Officer of Cognizant Digital Business. He was also founder of the Center for the Future of Work at Cognizant. And he’s a writer. His latest book is, What to do When Machines do Everything; we have copies and I hope people pick them up. It’s a terrific book full of interesting and provocative thoughts and observations. So Paul, welcome.

Roehrig: Thanks.

Petre: Sheila Lirio Marcelo is a founder. She’s a founder and CEO of, which is the largest platform on the internet for family care. It means elder care, it means child care, it means pet care, it means house care. It’s a platform for independent care workers to link up with jobs. So welcome, Sheila.

Sasan Goodarzi is the Executive Vice President at Intuit, and you know Intuit from TurboTax and QuickBooks. Intuit serves millions and millions of households and small businesses, and has deep insight into what those customers are up to, and they need to do that to keep their products current. So we’ll be tapping into some of that knowledge and insight from you. Welcome.

Diana Farrell, Diana had a long, distinguished career running the McKinsey Global Institute. She then did a stint at the Obama White House as a Deputy Director of the National Economic Council. And then came back to found her own institute at JPMorgan Chase, where she’s the founder and CEO.

And finally, Byron Auguste, who is our only walking, talking labor economist on stage, Oxford- trained. He had a long, distinguished career at McKinsey as well and wrote for the McKinsey Global Institute. He then also did a stint at the Obama White House as a Deputy Director of the National Economic Council, and now he runs a startup, which is called Opportunity@Work. And we’ll hear a little bit about that later.

But to get us going, and since we’re kind of in exploration mode this afternoon, I’m just going to ask one question and ask for a brief answer from each our five discussants. And then if there’s anybody on the floor that wants to throw in an answer to the same question, just raise your hand. We’re going to start the discussion immediately; we’re not going to wait until the end.

So here’s the question: Here we are at this moment of what the economists call full employment. Can you name a single thing that you find most puzzling about the nature of work in America and the job market today?

So I’m going to start with Paul.

Roehrig: The single most puzzling thing to me is there’s a zeitgeist of fear and concern. When you look back historically at every major shift in business and technology, there’s been an interregnum where people are afraid of the technology—either concerned or out and out afraid of new technology and what it’s going to mean. It could have been the Luddites, it could have been—you can go back thousands of years and find history. And I think we’re in that space right now. So the most puzzling thing for me, and we touched on in the book a little bit, is this sense of apprehension and concern. It feels very much like, you know, when you read the news, whether its political news or work news or economic or technology news, there’s a sense of things are not going really well. And so I’m puzzled and entranced by that context.

Petre: Got it. Why won’t people take some good news?

Roehrig: Well, you know, you don’t want to trivialize it and say, you know, everything’s great. But there is a sense of opportunity and possibility and I think that’s being turned down more than I think it should be.

Petre: Okay, great. Thank you. Sheila.

Marcelo: Hello, good afternoon. The most puzzling thing for me is actually how we undervalue care in this society, when it is the foundation and infrastructure that actually drives jobs and thus the economy. From a demand side, you know, it’s $320 billion dollars. It’s one of the fastest-growing job categories that supports all work. And if you’re wondering, in the audience, like, does it even pertain to me? So how many of you have kids? How many of you have been nieces and nephews? How many of you have been children?


So—and I often ask that question when I’m public speaking because people don’t think that the care relates to them. But the senior tsunami that’s coming of 10,000 born—I mean 10,000 that’s becoming 65 years or older, it’s upon us. And yet we don’t think about it. But yet on the supply side, we undervalue this infrastructure. I mean, we invest in roads in bridges, and yet what drove the economy in overall GDP growth was really female participation in the workplace in the last four years.

Petre: Okay, so we’ll put a marker on that. We’re going to come back to it.

Marcelo: So pretty puzzling for me.

Petre: Very good. Sasan.

Goodarzi: So just to humanize it, you’ve got folks today—they have a full-time job and they may do Uber, going and coming back to work. Or all they do all day long is do work, whether it’s, Upwork, DoorDash, their whole entire life is doing different gigs. And for me—and by the way, our entire United States system does not support these folks, in terms of how we do paychecks, insurance, benefits, everything. And the most puzzling thing to me, when you talk to these folks is on one end they absolutely love the freedom of being able to do what they want to do when they want to do it; on the other hand, they are extremely concerned about the lack of stability of pay. And that’s very puzzling to me because there’s a love-hate relationship that people have in this gig economy.

Petre: Interesting.

Farrell: So this is a hard question, because I think there are many topics and each one of the ones you’ve raised is good. But I think the one that I have not seen a good explanation, and I certainly have spent a lot of time trying to think hard about it—you alluded to it, Peter, in your 10 factors of the economy—but it is what has really been a 40-year structural decline in startups in America. And of course, this last recession, you know, made it terrible. But you just have to zoom out to say in 1978 the rate of new startups was about 15 percent of the base. That’s 800,000 jobs being created every year by people starting up new. Today, that number is 8 percent, fewer than 500 jobs. And you know you had a little bit of an up and tip, but it’s been a structural decline.

Petre: Yeah, we heard John Chambers talk pretty passionately about—

Marcelo: About this topic. And I thought he was very interesting, but I was not fully compelled by his explanations.

So I think really getting behind that at the same time that the exit rates have not changed, that creates a very clear problem for the structure of our economy.

Petre: Okay, great. Byron.

Auguste: Thanks. Well I think it’s not surprising that people feel afraid, because I think half of Americans certainly have been living through a very long and deep recession, I mean a multi-decade recession, effectively, in their household incomes. But what I do find surprising is the sense of fatalism among elites, if you will, across the board about the future of work. I mean, work is solving problems. And last I checked, we were not running out of problems. I mean, work is building things, it’s fixing things, it’s inventing, it’s caring, it’s entertaining, it’s educating. There’s so much work to do; there’s so much high-value work to do. I mean there’s so much high net present value investment to be done in infrastructure, in so many different phases of industry and invention. And there’s extremely high social value caring work to be done—which, by the way, I think has something to do with the entrepreneurship problem, when you have the disappearing fabric of care that provides the risk management underneath that. So I think we need to really look beyond—too much of our debate treats technology as if it’s a force of nature, and technology will do X or technology will do Y, or it will take this job or that job. And too much of it treats our institutions as static. Well, news flash, we are in charge of our institutions, we can change them. So if our institutions, whether it’s our capital markets or our labor markets, our education, are not doing what’s required to help us unlock the kind of investment in sort of the high value work we need to do, then we need to change them. And so that’s where I think I would like to focus the conversation.

Petre: Okay, great. Okay, so we’ve got zeitgeist of fear and concern. We’ve got the undervaluing of care. We’ve got the love/hate relationship about contingent work, do-it-yourself, cobble together a living. We’ve got the 40-year decline in startups. And we’ve got fatalism and passivity among elites about dealing with how to configure the workforce so it works better.

Now, anybody else want to throw one in?

Audience Member 1: I would like to pose a simple question. What would an ordinary work day look like in the year 2035?

Petre: Oh and that’s a good one. That’s one we may want to end on today. But that’s a really good question. Do you want to add one to the list?

Audience Member 2: [off mic] We’ve got all these technologies that are supposed to be making us more productive, and yet for the last decade plus, the productivity numbers have actually declined and are the lowest in a century.

Petre: Right. Okay. With all this automation, how come productivity’s not going up?

Okay well that’s a good place to start, actually, Paul. Because I know that you, in your book, would probably argue that we’re not automating and working on processes fast enough, there’s not enough investment going in and investment is what drives productivity.

Roehrig: Well it’s exactly right. The idea—so you’ve asked the question twice. I think it’s one of the right questions to ask in terms of why aren’t these numbers being recognized in terms of productivity improvement? And you know technologists talking about how great technology is in the shadow of Silicon Valley is not particularly surprising, right? But the question does come back to, okay, well where’s the impact? And I’m not sure that there’s not an impact or whether it’s—and you know maybe Byron can help us. I’m not sure that we’re measuring the right things in terms of technology and the productivity impact that I think is occurring.

The idea around automation and in fact AI over all, is that it is the next generation of productivity improvement. It is the system that can allow a higher degree of throughput in most business processes and customer experiences. And that is starting to, when you look at the evidence and what different companies are doing across different sectors, that is what’s beginning to show up. You’re seeing new commercial models, you’re seeing valuations change, and other proxy measures to productivity.

Petre: Great. So that’s a theme that’s come up a couple times already, is that we’re not valuing things and measuring things in a way that’s quite appropriate. Byron, you said something when we were talking on the phone that completely blew my mind. I’ve been obsessed with it ever since. You were speaking as a labor economist, and you said that the way that work is valued in labor economics is like the way that money and investing was valued in standard economics when the idea was that investors will always make the rational choice. Now classical economics had a revolution of behavioral economics come in with Kahneman and those guys, and it’s a lot more sophisticated now because it takes humans into account. But you said that that has not happened in labor economics. And it sounds to me like there’s a Nobel Prize just sitting out there waiting for somebody to pick it up.

Auguste: Right.

Petre: But could you explain what’s missing there?

Auguste: Yeah. The Nobel Prize will probably go to someone who’s working on a model right now, not sitting on a panel. But yeah.


Richard Thaler just won the Nobel Prize for behavioral economics applied mainly to investment and saving decisions, consumption decisions about the psychology of how people actually work and what they care about. And my point was that there’s an even more fundamental disconnect between the way people are modelled as workers in the standard economic model, that has not been disentangled. And I think we’re not going to really understand the situation we’re facing until we disentangle it.

To slightly oversimplify, but not by much, when you do the demand and supply, if you think of the labor market as a market, companies that want people to work, to build products, that’s the demand side, right? So they’re doing something useful. People, it turns out in this case, are the supply side, the way it’s modelled. So you offer your labor because you are paid. If you weren’t paid, you wouldn’t offer your labor. And specifically, the underlying assumption is that working, per se, is a negative. It’s a bad thing, right? But you do it in exchange for something. You give up your time; your leisure is what’s valuable—you give up your leisure time to work because you get money for it. That’s the microeconomics underlying all of our labor market models.

But from psychology we know, that in fact, having meaningful work is sort of a top three—it’s right at the base of Maslow’s hierarchy of needs, along with loving relationships, along with care and shelter, for someone’s actual utility and well-being. So how can it be that something that’s actually a fundamental input into someone’s well-being, that people have a deep demand for, in fact, meaningful work, is something that our entire intellectual infrastructure around labor markets and work treats as a negative that you only do in exchange for money? And until we disentangle that, I think we don’t much of a chance of really understanding what’s going on in the labor markets.

Petre: So what would be a meaningful policy change that would come out of changing this economic thinking?

Auguste: So if you took that very seriously, meaningful work is an input into people’s fundamental well-being and happiness and completeness as people, you would take very, very seriously policies that made less disruption of work, of working. Even if someone had to lead a particular job. So for example in Germany, they have job-sharing arrangements through the recession. In the United States it used to be over—the first five recessions after World War II, companies only laid off about a third of the number of people they would have needed to keep their profits stable. Now it’s 100 percent. So now it’s about 9–10 million more people that were laid off this last recession than if we had kept the same arrangements as the first five recessions after World War II in this country.

We’re not taking seriously the cost of disruption, not just to the macroeconomy, but to fundamental well-being and to fundamental skill sets. Because most of your skills we develop at work, not just at school. And what’s more, if you’re out of work for six months you’re considered damaged goods and it’s hard to get back in. If 10 million people get laid off at the same time, a lot of them are going to be out of work six months later. So we’re letting this drift happen that throws more and more people off the bus and puts them out of meaningful work. And we don’t need to do that. Our institutions don’t have to work that way. There are different ways we could do them. And we should change them on purpose.

Marcelo: Can I chime in on this?

Petre: Yes, please.

Marcelo: I also think—Byron mentioned, you know, the challenge around institutions, being a public company and the focus we have on short term and this EBITDA-operating income focus, that really is about reducing labor as much as we can to drive overall productivity. Instead if we were focused on long-term growth and investment, on creativity, value and what’s important to human beings, and we gave that long-term focus and allowed that growth enumerator to grow rather than the denominator to decrease, we would actually improve productivity in this company. It’s an interesting thing around our capitalist institutions, as well, for scale and where we find our monies has a value set that is also counter to overall improving a productivity that supports what’s important to humans.

Farrell: I think you’re right, Byron, that we would think very differently about this if we didn’t treat this as a supply issue but as a need. I think the best policy focus that that would bring to light would be if you look at the employment numbers and your subsegment, because it’s never useful to look at aggregate numbers, you start seeing that there are different issues that are just not getting attention. The biggest and most obvious one, related to what you worry about all the time, Sheila, is, you know, daycare for women. I mean there’s a reason that labor force participation has, you know, really become stagnant for women, is that there’s a structural problem. And if we sort of thought, “No, it’s really important for women to work, not just because they’re suppliers of work, but because it’s part of who they are and their identity,” we would solve that problem.

You look at our criminal laws and the ways in which we’ve taken kids in the inner city who, you know, had some drug run-in or another and labelled them criminals and basically kept them out of the workforce for 10, 15 years. We would solve that. We have an opioid crisis that is a health issue that is keeping a lot of people out of the workforce. We would solve it. We wouldn’t just call it an emergency and then turn away and pretend nothing was happening. And I think that we aren’t being creative enough about understanding where are the pockets of real problems with either labor force participation or employment and policies that are going to be very targeted to bringing people back into the fold and also back into the workforce. Because that’s an issue. Our unemployment rate is 4 percent but our underemployment rate and people who have been discouraged is much higher. And we can solve this if we had a very different mind-set towards it.

Petre: Well my question is that you have a microscope—I mean, even to survive as a company you have to understand how small businesses and households are managing. And what we’re hearing is that people are cobbling together jobs, their retirement security has gone to hell. I mean, a lot of factors have shifted among your customers. And you guys not only can see how they use the software and what they respond to in terms of the tools you provide, but the research that you’ve supported as a company has really looked at household patterns in very interesting ways. So I wonder if you could speak to that a little bit, like what you’ve uncovered?

Goodarzi: Sure, I’ll start with actually what we’re experiencing as a company and then I’ll hit on the customer piece in a moment. So we have 8,000 full-time employees, but at our peak during any year we hit almost 15,000. And the 15,000, they’re all contingent workers, to be able to support as we go through what we call busy season and globally. And so what that’s actually taught us is our payroll systems, how we engage them—one of the things that we do is we make sure that we pay our contractors net 10 days, versus most folks pay them the normal 45 days. These folks that are contingent workers, they have a hard time surviving. So as a company, because we hire so many of the contingent workers, we’re actually learning how to operate differently to make sure that we can help them with their well-being. And frankly, just survival, right?

Back to the question, though, that you asked around our customers. You know what hasn’t changed is that there’s those that are self-employed, whether it’s the pool guy or the pool girl, or the one-woman, one-man shops. That hasn’t changed. Those are self-employed. But what has changed is what I mentioned earlier. You run into a lot of folks—around the globe, by the way, it’s not just in the U.S.—where they have a full-time job but they’re either doing a side gig as a designer, through Upwork or they’re Lyft or Uber drivers, and that changes the tenor of how they even think about managing their personal versus business expenses. Do I wait until tax time to pay all my taxes? Or when I get all this income that’s side income, do I need to pay taxes now?

Petre: Right. Not to mention deductions and pass-throughs and—

Goodarzi: That’s exactly right.

Petre: —it’s a growth business, right?

Goodarzi: That’s exactly right. The gentleman over there asked what will the world look like in 2035? Your guess is as good as mine, but we think one part or a big population will be folks that are just doing a bunch of different gig works. And just in terms of how you find work, the stability of income, how you manage your personal versus business deductions, how much you put to the side, how you connect with insurance and benefits, is just a massive change that we’re seeing. And so we think about that in terms of then how do we create a platform that helps put more money in the pocket of these workers? How do we eliminate the drudgery work and drive automation so they can focus on what their passionate about? And so those are some of the key insights, which I don’t think is any different than what the panel’s actually already shared.

Petre: Yeah, but you’re raising a question that’s very important. And Sheila, I would love to hear you talk about this. You have an initiative going in your company to help bring workers who are on-demand, essentially care workers.What’s the average? $9 dollars an hour for a care worker? And to change that so that they have some support systems that are like the support systems that a full-time job used to provide when you worked for a company. Can you—

Marcelo: Sure, I mean, it’s interesting. I know the drivers get a lot of attention in the gig economy but we have 12 million caregivers on the platform thus far on We felt like it was our social responsibility because it’s such an undervalued part of society. So we provide household payroll, so that they can get access to social security and other benefits to families. We also provide worker’s compensation in all 50 states. It took us four years to build that, because families are so confused to say, “I’m hiring someone for my home. What’s worker’s comp? How do I even do this?” Provide access to healthcare through a partnership we did with Stride. And then we launched, for the first time, the first-ever pooled portable benefits platform. So a little bit of conscious capitalism. So for every transaction, we actually put 2 percent of’s fees towards the caregiver as a benefit, and then another 2 percent that we ask the family to give. And that’s up to $600 dollars a year.

But that, if you think about a corporate benefit, is equivalent to, annually, a pharmacy benefit. And they can use that money towards premiums and currently roll it over even to a subsequent year. So we’re trying to create this sandbox role modeling. Byron and I have spoken on panels together and finding ways to work—how do you create this social net that is so necessary? And then, on top of that, it’s not just even benefits. How do they get access to training and re-training so that they have career pathways? That caregiving is just a stepping stone so that they can actually go into home health aide, into nursing, because there’s a shortage there too in the country. There’s ways in which we can really use the power of the platforms for reach. Because the problem with a lot of nonprofits and some of the for-profits is the challenge of distribution. So as a consumer platform, we feel it’s our responsibility to actually do that advocacy and reach and use the power of technology, without charging, specifically, because we already monetize in a different way. How do we create a social net specifically for caregivers?

Petre: So before we go away from that, among the families who use to use caregivers, what’s been the uptake of these things you’re offering? So in other words, if I hire a housekeeper through you, I get an option of whether to pay minimum wage or I can just talk to that housekeeper offline and say, “You know, really, it’s going to be $9 dollars an hour. Do you want the job?” So can you talk about what the mix is?

Marcelo: We haven’t because we’re public, but what we’ve done is part of our initiative, of what we call Care 3.0. For a long time, we built the matching platforms, so millions of matches and a high efficacy. It’s a 90 percent match rate on Now what we’re doing is building what we call beyond the match, which is Care 3.0. Once you’ve matched on, what kinds of things could we be providing for you, additional services? So whether you hired on Care or not. And that’s the fastest way we can actually get to advocacy. So that we’re in the relationship, asking them to really treat the caregiver in the right way.

Petre: Interesting. Byron.

Auguste: You know, just listening, it always feels to me like the gig economy worker, the contingent worker, should better be thought of almost like a canary in the coal mine for the entire workforce, in the sense that the challenges that contingent workers face are actually challenges that many other workers who are not contingent face. And I think we underestimate the degree to which our institutions are problematic, even for many that are in the full paid workforce. And I just think it’s just easier to spot in the contingent workforce.

So for example, if you talk about the challenge of scheduling these different gigs and how do you manage all those? We have almost 20 million American workers that are technically full-time that are on scheduling platforms for their companies, to optimize their company’s convenience. And that they can, on four hours’ notice, after they’ve arranged childcare and they’ve not gone to the community college class that would have helped them get to a better place. They just don’t have the shift, right? They find out at the last minute. And that happens all over, to 20 million people. So that’s a good example there.

You talk about the training and the lack of training in the gig economy, well, I’ll tell you, it used to be that it was pretty standard to train people across the board, and companies still do, but right now the difference between if you start, coming out of, say, a selective university, into entry-level professional, managerial or technical, then your company puts a lot of training behind you with the intention of having you move up. But if you start on the retail floor, if you start in the call center, if you start in the warehouse, you’re probably being trained for safety compliance and efficiency. Obviously, there are exceptions, there are companies that still do a lot of that training, but it’s a real issue.

And then even the tax challenges, of dealing with the tax code being biased in some ways against contingent workers, well the tax code is biased against work, period, right? If you try to accomplish the same—if a company has a new product or service and they try to deliver it, let’s say by deploying lean and sort of training their workers to do it and, you know, empowered teams, cross-training, and they—same cash-in, cash-out, they do it with machines and software instead and take out workers, they have a huge tax benefit. Even though it’s the same cash-ins, same cash-outs, same economic efficiency. If you do it by training people and employing them, you will have a lower accounting earnings, you will have a lower PE ratio, you will have lower executive bonuses. So we have absolutely biased our institutions, our major institutions of capitalism, against work across the board. Not just contingent work. Any work, by humans, that is.


Just humans. The rest is fine.

Petre: I see a question here.

Neuberger: Hi, thank you, this a great discussion. I’m Lisa Neuberger from Accenture and the topic you just raised, Byron, is something that I was going to raise as a puzzling point. So what puzzles me is about re-skilling, up-skilling, re-training. So you see the big companies like Accenture, like GE, like Daimler, like Toyota, investing heavily in massive re-training, re-skilling initiatives right at this moment. But who’s re-skilling the workforce who works in small and medium-sized enterprises and the gig economy? So I would love your thoughts on where are the green shoots? Penny Pritzker spoke about the green shoots that are coming up in this space of re-training and where are we seeing that in the workforce of SMEs and the gig economy and what kind of solutions do you see?

Petre: Can I interrupt to throw in another gloomy factor? Byron, you were talking about this earlier. At the same time these pressures are on the workforce, employers—and Paul, I don’t know if you see this among the people you work with—seem to be getting more and more conservative and risk-averse in hiring. There’s what’s called ‘up-credentialing,’ where you’ll advertise a job and you’ll say, “Only BAs accepted,” where the job doesn’t require a BA. Or the slightest smudge on a worker’s record—you know, you can collect 2,000 resumes where you used to be able to collect 40. And of those 2,000, you apply sort of the most brutal filters. So you end up with somebody who’s probably overqualified for the actual job and stiffing people who really could use the work, and would be just right for the job. So there’s a distortion that seems to be happening on the employers side and I can’t explain it. I have no direct experience with it. But I was hoping someone here could shed light on that.

Roehrig: Yeah, so I’ll try a little bit. So there’s a couple of themes going around Cognizant employs 260,000 people or so. And so we have and other large companies like GE, Accenture—there are engines for collecting, you know, information for training. I mean these large companies, it’s incumbent on them—on all of us—to make sure that we are taking the appropriate steps to get people skilled for digital economy jobs. To not do that—there’s an economic imperative. There’s an innovation imperative and there’s an ethical imperative. And most successful big companies that I’ve worked—that we work with, there’s not anybody that’s just gone, “Nope, not gonna do it. Training is a bad idea.” I’ve never heard that, not once.

But to another point that’s come up, and I think, to the notion of the tax code creates a downward pressure on humans. I think that’s true if we assume that throughput is the same. The idea of AI and automation being applied—and this is happening across multiple sectors, whether it’s in the industrial internet or technology services or banking—or healthcare—is that if you lower cost, the actual cost of the means of production, the market will ensure that there’s more throughput. So the dystopian future that is painted frequently is that demand is fixed. And that’s not right, right? That’s never been right. Like you can go back to Henry Ford, there are plenty of historical examples. But the idea that if the cost goes down, you can—and we’re doing work in a hospital, actually, where we help improve throughput, and now they’re serving more patients. The cost did go down, but nobody left. In fact, they’re growing, because it’s more efficient, it’s more effective. And I think that’s part of this equation. We can get very easily especially in kind of the post-lunch sugar crash that we’re all kind of living through right now, it’s like, “Oh God, the robots are coming.” We’re all just “At least the view is nice.” But the point of this is, though, it’s not necessarily dystopian and we’re in Silicon Valley, so I’m fully cognizant—pun intended—on where we are. They’re saying, “Technology is going to make everything perfect forever.” And then, you know, you’ve got other folks that are raising important points, Elon Musk and Stephen Hawking are raising important questions, but they’re highly dystopian. It’s like we’re going to—let’s just build a fire and end it.

But then there’s a pragmatic middle here that is much more plausible and historically justified.

Petre: Okay, so let’s stay with that though. But something Diana said at the beginning is you have teased apart the different segments.

Roehrig: Right, absolutely right.

Petre: And Byron, you made the point that it used to be that the labor market worked for 80 percent of the workforce. And now it only seems to work well for 50 percent of the workforce. And I didn’t quite understand what you meant by that.

Auguste: Well, yeah, you can debate the percentages—and I know we haven’t gotten to your question yet, but I’m sure we will—

Petre: But we’re circling it.

Auguste: We’ll circle around to the training. But I just wanted to respond to Paul because the point I was making about the way our system is biased against labor—I mean partly the tax code—but actually more the accounting code, even more so than the tax system. So it’s actually found to be accounting, which interestingly is not a government thing, per se, but it’s a nonprofit. It’s an industry standard. I’m not saying that there’s not going to be enough demand. I’m actually quite optimistic on that point. To my earlier point, there is so much work to do. There’s a tremendous amount to do. I mean, we talked about caring, we haven’t talked much about infrastructure, we haven’t talked much about travel and tourism and new experiences, as well as medicine and new cures. I mean, there’s so much work to do.

So I am not at all pessimistic about the demand side of things, like the aggregate demand, the absolute demand for new things for humans. But it’s not just that the market decides. This is mediated very much through institutions. So my point about, in general technology, I think, tends to be neutral. You can use it in all sorts of ways. You know a hammer can smash your head or build a house. It just depends what you do with it. AI actually, and what you’re seeing right now in the next 20 years, classification, all that. Actually, I don’t think it’s neutral. I think it wants to augment people. It’s mainly taking off in ways that are part of jobs, not entire jobs. And yet, if you look at the way it will flow through our business models, our capital markets and the like, it’s very likely to be used in most places as just another form of automation. Where it really could be a massive form of augmentation. Where you will see it used in augmentation is in healthcare. Why? Not because the market decides. But because the guilds are strong. You can’t replace doctors. The AMA won’t allow you sort of change that. So therefore, we’ll use it to enhance what doctors can do, their effectiveness and so forth. But we could use it to enhance what nurse practitioners could do, what home healthcare aides could do. Oops, except it bounces up against the scope of practice laws that says a home healthcare aide can only do so much, no matter if she’s augmented by technology or not.

So it’s not a matter of just the markets. It’s very much a matter of our laws and institutions and our political economy that determines how technology flows through and how it affects who’s job. So I just wanted to—like we can’t expect—markets don’t descend from heaven, the clouds part. They’re human institutions. We create them. And we need to create them better in the way that they work for more people.

Marcelo: Can I address the training Lisa asked? Is that okay?

Petre: Please.

Marcelo: So I actually think it’s got to be both sides on the supply and demand side. In terms of the gig economy, the way we think about it is we’re launching the Care Institute, which is a separate 501(c)(3), partner to the AARP, rolling it out later this year, that really trains caregivers across the country. Hoping to partner with Byron, so a lot of work to do there. But then there’s also, what are we doing for families? Because if you’re solving the supply issue, there’s a lot of low-income workers. Byron touched on retail workers, whether they’re at banks, McDonald’s. You know, if we’re only servicing the white-collar workers then we’re not really going to support overall productivity.

But how can they afford that? So affordable care. What we’re doing is actually training as well, it’s called HomeStart, training family day care providers, low-income immigrant women, to run their own business, so they can serve the local community, especially when it comes to backup care. It’s really difficult scheduling of retail workers; they can just call a local community, a family day care, to serve their needs. So there’s a lot of solutions out there. And how do we take private-public partnerships, using the power of technology, to actually solve some of these things? And a lot of these programs have been in existence, but you can’t actually fill a family day care. And we realize the power of is we train these women; we spend a lot of dollars training them. But a lot of nonprofits are just fed up. You know, because the women decide after a year they’re going to leave that profession because they can’t fill their local day care. And so where can we play as technology to provide a solution? And again, it’s that combination of HI and AI that has to come together on both sides.

Goodarzi: Can I add to this? Because actually, your question is a really good one, where I’ll provide a part perspective and I think there’s other elements that will get answered over time. Because I don’t think we know. As I mentioned earlier, we have 8,000 full-time employees, at our peak. Every year, we add another 7,000. And they’re all contingent workers. And what we do is we bring them in and we train them. We train them to deliver for our customers—this is in our customer success organization. At the end, we do the exit interviews with every single one of them. One, what was their experience with us? What can we do better? And then provide them input in terms of where they can go get more training. And specifically suggest where they can get it. So I think we, collectively, can play an enormous part in that. We do the same when we hire contingent designers, product managers, when we have peaks, engineers. You want that to be core, but we also provide training and also suggest where they get training. That’s the perspective that, at least we know today. There’s a huge perspective that I think goes unanswered; you have to be self-directed to get your own training. All these folks that are doing different gigs—how are they going through and getting themselves skilled and trained to do a great job? That’s something that we have to actually figure out collectively over time. But I think there’s a lot we can do to provide that training, like you’re doing, which is wonderful, to skill the workforce.

Auguste: This is the main thing that I’m working on now with Opportunity@Work, which we founded two years ago. In the White House, this was one of my domain areas, the workforce training. And I would agree with everything we’ve heard here, that first of all, a lot of companies do a very good job. Certainly, many companies do an excellent job with their own workforce, and many companies try to extend that a bit into the community in terms of partnerships and the like, and there are many, many that are effective. Cognizant and its peers, for example, in IT services are tremendously effective at training people.

So to the point that Peter mentioned, so I think there’s really three things. That if you want to talk about how you take these things to scale, beyond, you know, a few apprenticeships here and there. I mean, there are so many good things, that if you add it all up, there’s no way they can get to more than, say 15 percent of the need, every good thing, in every way that it could work, that you see out there. We need to have a different model that can take the good and scale it. And I think there’s three things that matter here: number one, you have to have more people have access to those career path jobs. And this is to your point, if you say that you have to have a bachelor’s degree to do a job, that you didn’t learn anything in your bachelor’s degree to do that job, you’re just using it as a very crude signal. If we are going to equate someone’s skills, someone’s potential, to how much educational attainment they have, then a society where educational attainment of that sort requires social capital, it requires, savings, it requires money. So it’s all sorts of things that don’t correlate to talent.

We are defining ourselves into a skills gap. I mean, at most, a really insane example: 20 percent of administrative assistants have bachelor’s degrees but two-thirds of the new job postings for administrative assistants require a bachelor’s degree to be considered. So 80 percent of today’s admin assistants can’t apply for their own job, for two-thirds of the new jobs in their field. But you know what else for administrative assistants? In Opportunity@Work, we’re seeing a lot of admin assistants from small businesses who don’t have an IT department; [in] these businesses, they are the IT department.


They’ve mastered these things. But now to apply for a career path job in IT. You won’t get into the interview. Because you’ll be screened out through the applicant tracking system before any human has seen you, before anyone has assessed what you can do.

So the first thing is you have to have a system where it shows what you can do and our jobs will look for you. And that’s what we’re doing first in Opportunity@Work. We have a network of, right now, about 1,400 employers, and we’re starting with IT jobs. We have three up on TechHire Careers and it’s moving. But we work with the employers to develop these assessments. We have a network of interviewers who are volunteers, who are the same person who might interview you for a job at Amazon, but now is giving you a 30-minute mock interview where you get in the pool to be hired if you get through. And if you don’t, you get feedback specifically on what you should work on, which you can’t get in an actual job interview if you could get one. So that’s critical.

And then second, you have this network that now you can start to define: Here are the skills you need to be able to do these jobs, so that training providers of all sorts now have a target to shoot at. Businesses say that training and education don’t know what skills we need. How would you know what skills you need if you don’t tell them? You tell your suppliers your specifications, right? These are like the suppliers of your talent, you don’t tell them your specifications. But you can do that with the network. And then third, the kind of innovation that we talk about at Techonomy, which is applying right now in the talent tech, at the top of the market, to create these super empowered free agents. And for companies to sort of plug in your people like widgets. We need that technology for individuals, all the way down to the most marginal individual, to be agents and actually support them, negotiate on their behalf, figure out how to plug in their training.

And we’re just starting to do that with that network, because you start to see what skills you need to take that next step. So companies together—companies individually, like you’ve mentioned with Intuit, can make a difference. But companies together can make a massive difference. Because if the millions of people who apply for jobs at big companies and are screened out and get this electronic form letter, if they also got a link to a place, “Well hey, you don’t qualify based on your pedigree. But try here, and see if you can qualify based on your performance If there’s always a lane where you can be screened in based on performance, not screened out based on pedigree, then we can get started for people turning their effort and their learning into earning.


Petre: Okay, there’s an issue I promise we would get back to. And that is the mystery—it’s a job-creation mystery, and it’s the mystery of business formation. And Diana, I know you’ve flagged that as the biggest puzzle.

Farrell: I could touch on that, but I did notice there were at least three people who were trying to get in the whole time.

Petre: Okay, so let’s harvest some questions.

Farrell: And then I’ll try to answer your question in a minute.

Audience Member 3: Not at all. This is an awesome panel. No apologies please. I actually did have a question that was primarily for Diana and for Paul, though I would love to hear the entire panel’s opinion on it. Clearly we’re experiencing the law of unintended consequences. I remember four or five years conferences on the future of work where you had companies like LinkedIn and WeWork very confidently predicting exactly what to expect. And this entire panel is now about the opposite. So my question is, where are the biggest blind spots we have about the future? Where is it most important to be tracking the data and looking for emerging trends? Assuming that we are actually flying through the fog right now. What are those signs, those indicators, those KPIs? Where does data and analysis most help us chart the future, like one day at a time? Where are the blind spots? What should be looking at?

Roehrig: I think you should go and then I’ll agree with you.


Petre: We’ve got varied data in terms of people here.

Audience Member 4: Yeah, so what I’m worried about is sort of the downward trend and what you could think of as the bottom and the middle class. And there are three forces you talked about. And I’m just worried about what we do about it, because it feels to me, if I listen to the story you’re describing, it will continue. So one of them is more and more people moving to care work, but we actually don’t value it. And therefore more and more people doing important stuff but actually not getting paid very well.

The second is what you raised, which is actually because we overqualify people for jobs, actually people are getting paid less than they should, given the thing they’ve gotten, and therefore there’s a downward progress. And then the third is, I think Diana, the point you’re raising, which is we actually don’t—we aren’t creating the companies that actually create the work that then creates the next generation of wealth. And so I think the worry I have that’s dystopian, the worry I have is that a larger and larger percentage of the population is getting to a point where their lives are less and less attractive, for those three forces. So I’m just curious about your thoughts about how to respond to that challenge. And I think I would like your answer first, I guess, because creating new businesses is a huge piece of it.

Petre: That is a very eloquent question, thank you. Okay, I’ll take one more.

Audience Member 5: Just a quick add to that, right here. We’ve talked a lot about outside the company, but inside the firm, we do a lot of development for large firms. And engagement scores are way down. And when you talk to people, it’s because, “Well I get my medical here and you know I’ve been here for 15 years. If I can just hold on for another five, even though it’s shit work, it’s better than going into the gig economy, because I heard that sucks.”

So I think there’s another side, which is the corporate bureaucracy and hierarchy that we live in. There’s some that navigate to the top and are these super human, float around and do super cool stuff, and then there’s like the unwashed middle that’s just droning around, hoping just to get by. And I think there’s a lot of cognitive surplus that needs to be unlocked in there too. So it seems there’s a double story here, picking off what Blair’s saying. Inside the corporate bureaucracy, there’s kind of latent potential locked up. In the gig economy, there’s this kind of fluctuation and variability, and if you ask the people in the corporation, they would just love to go and do their startup thing, to Diana’s point, but they don’t like the volatility. And for those who are in the volatility, we’re trying to create stability. So I’m back to Paul’s magic middle. There’s probably something in the middle where both sides could see benefit in creating some kind of marketplace.

Petre: Okay, we have one more question.

Bolles: My name’s Gary Bolles, I’m the Chair for the Future Work at Singularity University and of the startup B-Parachute, built off the work of What Color is Your Parachute?

Petre: So this is your wheelhouse, right?

Bolles: It’s my favorite topic. I just want to play off something Byron was saying. So not only is our legislative system biased against work, but the Silicon Valley innovation machine, to some extent, is biased against it as well. You’ve got venture capitalists funding innovation, which is basically to go automate tasks that humans are performing because that’s the money that’s being spent today. But if we don’t take that same amount of innovation that’s being focused on replacing work and focus it on augmenting and upskilling people, the results are kind of inevitable. So we have this massive machine that could be turned towards providing a lot of the same kind of upskilling and rapid training and even embedding training within work. But the machine’s not calibrated toward that, so. Thoughts on how to change that?

Petre: So Diana.

Farrell: Okay, in one minute or less. There were so many good questions and we won’t handle them all. But let’s come back to this startup issue. I don’t think we have good explanations, and I think if we spent more time thinking about that, we would unlock a lot of things that would help unleash many dimensions of the problem. One of the things that people like to say, and I think John Chambers this morning was saying some of this. It’s this sort of creeping regulation that has prohibited entrepreneurship. I don’t think he used those words, but I’m very skeptical of that argument. Because we have natural experiments across the country where some states are highly regulated, some states are relatively lowly regulated. And you think of California, it’s one of the most highly regulated states, and you actually have higher startup rates here. So I’m sort of skeptical of the creeping regulation argument. I think there’s the Walmartization argument that killed the moms-and-pops, and maybe now the Amazonification.

Petre: Platforms absorbing the—

Farrell: Just scale. Scale wiping out sort of lots of mom-and-pops that were sub-scale but that exist in a different plane. And there’s no doubt there’s been a lot of that. But it’s also true that those are places that have given rise to startups as suppliers and others, so there’s not a foregone conclusion that that model would have to end up with the decimation of startups, etcetera. There’s this psychological argument that says millennials are just not risk-takers and baby boomers were, or that we’re just an older society. And again, I could piece that apart if you had the time to work through it.

Petre: Your Institute’s founded by a bank. What about credit?

Farrell: And I would say that that’s part of the regulatory argument, that says what businesses lack is credit, and that’s a result of the kind of capital weights that are put on providing risk to high-risk companies. You can blame the banks or you can blame the regulation, but you’re back to kind of a story of regulation.

And then there’s another maybe more optimistic argument that says, “No, what’s happened is that the marginal value to an entrepreneur today of going to start up something on their own versus going to a large company that is increasingly allowing entrepreneurship within those companies, is what’s changed.” So that if you were modeling the ways in which JPMorgan Chase is funding like all these entrepreneurs who are figuring out what to do with blockchain technology, within a bank, for example, in another world they might have been outside. So I don’t know. That’s the latest new theory I’ve seen. I think there’s some value to that.

But I think that only applies to the highly educated who would be attracted by these companies and not the broad range of people you were describing, Byron, who are really entrepreneurial. And they’ll figure out how to do the IT system even though they didn’t have IT training, who also are attaching themselves to large companies whether or not they’re getting that entrepreneurial spirit. And I do think that if we had a different view about it, that this is as much more of a beneficial thing to our country, even if it required subsidizing capital, even if it required subsidizing labor while people are in those jobs. We would begin to unlock that, which I think is going to be really essential. Even today, small businesses are nearly 50 percent of employment. So it’s a big issue.

Petre: Our time clock up here is failing but I have one here.


We’re about out of time. So let me just ask a quick closing question of all of you. It’s a takeaway question. If there was one thing that you would like to see business do or government do or individuals do to improve these issues that we’re talking about, and get employment, the job market, into a better place, what would the assignment be?

Auguste: So I just said what I thought business should do, and that is sort of joining what Opportunity@Work is doing, almost a talent alliance. So I’m going to take my minute to answer these questions, if you don’t mind.

Petre: Okay, no, no, no. I want to let people answer that question, then we’ll come back to that. Okay?

Auguste: Okay, great.

Farrell: One is the one thing we need to redefine the value of work? Not just the care question that you raised, Sheila, but the ways in which we bias the entrepreneurship that comes from pulling together independent work and otherwise. So I think a redefinition of our policy structure around what work is—

Petre: That’s a policy answer. But it has to be deep and thought through.

Farrell: I think so.

Goodarzi: You know I’ll focus my comments on just, again, those who are self-employed, which includes the one-man or one-woman shop and/or those who are doing a number of different gigs. And I would just say that our entire system, globally, is wired for those that work at companies that get paychecks.

Petre: It’s the W-2 people, not the 1099 people.

Goodarzi: That’s right. And we should think about the 1099 folks, from how they get paid to the training that’s provided, to insurance, to benefits. And think about how we focus on putting more money in their pocket, eliminating work and drudgery, and eliminating the fear that they have so they can thrive in that environment. Because that future is here. And it’s only going to expand.

Petre: That’s great. Thank you. Sheila?

Marcelo: I think building on valuing work and also valuing 1099s, certainly I would like to propose valuing caregivers, but you guys already know I represent that advocacy. But it’s actually asking you as individuals. You are incredibly talented, highly educated, have a ton of resources. And the fact is you can be a force for systemic change, to challenge the institutions, to challenge your definitions, to challenge how we do all these things, and the approach of what technology is going to happen to humanity. And so I think we all have an individual responsibility to use that purpose-driven creativity that is so human in all of us, to make a meaningful difference.

Petre: Paul.

Roehrig: It’s almost the same thing that you just said. We talked to companies all over the world, policymakers—and to boil it down to one thing, one recommendation that I would have based on all the research and everything that we’ve been able to look at, is don’t short human imagination. It’s been a bad bet throughout history; it’s a bad bet now and it will continue to be a bad bet. So whether you’re on the dystopian tribe or the utopian tribe, there is a pragmatic middle there. And the ticket to entry to that way ahead, is don’t short human imagination. To go on the one side, it’s like you would over bet on the obligation of the nobles to sort everything out, and on the other hand you’ve got revolution in the streets. The middle is possible, is plausible, and it’s up to us to make it happen. But it will never happen if we short human imagination.

Auguste: Well if it’s a policy recommendation, though, it’s really putting meaningful work first. So I agree with Diana but I think if you look at the actual regulation and then practices, business practices that we don’t call regulation but are just as meaningful restrictions on people’s lives. When you say that an administrative assistant who’s been doing the job for six years has to quit and go back to school for four years in order to maybe get the same job later, that feels like a regulation to that person. That person is stuck as a result. And we say California is regulated, but California does not have non-compete agreements and other states do. So the dynamism of California might be because high taxes don’t matter as much as a fluid labor market and California has a more fluid labor market for talent because of those things.

Petre: Obviously, this debate will go on and to the Techonomies—I want to thank everyone here, Paul, Sheila, Sasan, Diana and Byron. And thank you all for your questions. It’s been great.