What Happens to the Social Safety Net In a World of Automation and AI?


NewCo Shift Forum

A conversation around the future of labor

Left: Maya Rockeymoore, Ph.D. Founder, President & CEO, Center for Global Policy Solutions. Center: Stephen DeWitt CEO, Work Market Right: Moderator, Fred Wilson Managing Partner, USV

As I was planning the overall narrative arc of last month’s Shift Forum, it became quite clear we needed to address the question of how our society manages its workforce — in particular, the role technology, AI, and automation plays, as well as the massive role of government policy. I’m not a huge fan of panels, per se, but small ones run by nimble moderators can really bring a topic to life. As I was pondering the best way to do just that, I spoke to Fred Wilson, co-founder of Union Square Ventures. Turns out Fred was deeply interested in the topic, and volunteered to moderate the discussion. I’m certainly glad he did — his panel featured Maya Rockeymoore, Ph.D. Founder, President & CEO, Center for Global Policy Solutions, and Stephen DeWitt CEO, Work Market. Rockeymoore is an expert in labor policy, and DeWitt has pioneered a new approach to labor through his company’s Work Market platform. Below is the conversation in both video and as a text transcript, edited for clarity.

Fred Wilson (FW): We’ve been dancing around this issue all morning long — We get to actually have the headline. We’re going to talk about the future of work. Everyone else was talking about it too, but we got that headline.

It’s an interesting discussion because I’m going to be joined by two people who sit, in my mind, on the opposite side of the spectrum. First is Stephen DeWitt, who is the CEO of a company that I get to work with regularly, called Work Market, and they help large enterprises manage their workforces in an on-demand economy. Steve will talk a little bit about what that means.

Then we have Maya Rockeymoore, who is the Founder, President, and CEO of the Center for Global Policy Solutions. Maya thinks about, and helps her clients think about, where all this is going, and what are the policy implications, and what are the policy solutions that we need to put in place to deal with it.

I think it’s two very different perspectives but they both are seeing the issue front and center. I’ll start with maybe a provocative question. I wrote a blog post this morning about this panel, and I got an email from somebody at Harvard saying that the on-demand work force is basically a race to the bottom, and anybody who suggests otherwise is just being intellectually dishonest. I’m going to ask you both that question. Stephen, how do you react to that statement?

Stephen DeWitt (SD): I’m definitely not emotionally dishonest, but I disagree. The nature of our workforce is changing, the traditional definition of labor that supports an enterprise operating model has evolved. No longer is the talent, the skills, the passion, and the creativity to drive enterprise businesses — it’s no longer simply inside of your four walls.

Modern enterprises will leverage technology, engagement tools, big data, connected society to forge relationships with skilled labor, skilled talent. They may choose to engage with them in a traditional W-2 sense or they may choose to create operational clouds of independent workers. They may have suppliers, agencies, other partners etc.

What enterprises are trying to do, we look at this in three very simple ways. First, the enterprise has to get automated. Enterprise processes, business processes, the way you do things are evolved as a result of technology. When spreadsheets came out in the 80s, that displaced the old way of doing accounting. As database technology came out, that changed the way we did filing. We’ve all been part of this rapid evolution. Automation inside of the enterprise is a big deal.

The second aspect is the support for nontraditional arrangements with labor. People are changing their lifestyles. It’s hard for any of us to predict what life is going to be like 30 or 40 years from now, but we know this, our population is getting older. What’s going to be the engagement model with really smart people that have decades of experience, that are 70 years old? That’s not going to be a traditional arrangement. As you deal with all of these nontraditional arrangements, how are you going to manage that, administrate it, and do it in a way that drives the competitiveness of your business?

Then the third and final aspect of this is the worker themselves. We are in a connected society. McKinsey put a piece out that by 2030 there will be 3.2 billion skilled workers on this planet all connected. The only thing that separates people with skills and passions and desires and the desire to work is search.

FW: Maya, what do you think? Are we in a race to zero?

Maya Rockeymoore (WR): Actually the gig economy and the nature of part-time work can actually be a pipeline to poverty. You all have heard the whole notion of the retirement crisis in the United States, and it’s important to understand how that actually connects to temporary work, part-time work, gig work.

Approximately, 50 percent of US workers don’t get access to tax-preferred retirement vehicles through their jobs. As a result, as they age and go through life, they have nothing except for Social Security to rely on once they reach retirement. This is the same except perhaps worse, for gig economy folks, because they are not necessarily putting into the Social Security system at all.

What you have is a situation where people are doing gig work, short-term work where they are getting paid but they’re not getting retirement benefits tax-deferred retirement benefits and they’re not contributing to the Social Security system. That is a recipe for disaster, folks. Meaning that is a recipe for poverty in old age.

FW: As I predicted we have two different points of view up here. Maya, you mentioned, while we were waiting to go on, that you’ve been working on a study estimating what the job losses will be as we move to fully automated vehicles?

MR: I run a think tank at Washington DC and we are about to put out, so you are getting the first taste of what this report is about, what the impact of autonomous vehicles will be on jobs in the United States.

The estimate is there will be approximately four million jobs lost. That it will have a disproportionate impact on workers of color, and workers from certain states like Wyoming, Indiana, Idaho, Mississippi, and some other places that have a disproportionate reliance on driving jobs.

We need to be concerned about this, because there seems to be no coordination between policy leaders and the tech sector with regards to the rollout of this technology. We predict that the devastation could be quite significant, especially if the rollout of autonomous, driverless vehicles is done suddenly instead of over a period of time.

Policymakers and policy leaders should be actually a part of the conversation about how to put in place mechanisms and solutions to actually mitigate the damage that can be done to the US economy as a result of the new technology.

As we know AI itself, the projections are that over the next 20 years, AI can actually result in the loss of approximately 76 million jobs nationwide, and so we have a lot to be concerned about. There are some people who argue that, we’ve seen in the past when the cotton gin went away, people were absorbed into other sectors.

One cautionary tale in this regard and with regards to that response, is that technology has become so efficient and we’re iterating, ideating and innovating at a rapid pace, perhaps more so, than we are doing in terms of job creation.

There are certain policies that might help to spur more job creation, but I don’t think that we’re coordinating in a way that will allow people to develop the skills, and the opportunities necessary to create jobs as fast as they’ll be losing them as a result of technology. [applause]

FW: Steve you ended, when I cut you off earlier, with the notion that search is the only thing sitting between somebody who has worked and somebody who has a skill.

We just have this notion that AI could destroy jobs. Is it also true that AI could facilitate work that is otherwise not going to happen? Because your company has a sophisticated AI system, are you going to be able to connect two people to get something done that otherwise would never have connected?

SD: No question about it. Not only no question about it, but also across the universe of suppliers of goods and services that could transform your business.

A couple of points, I think there’s some really good points that have been covered here and been covered in this conference over the last couple of days. The nature of work is changing, and technology, we’re not going to put the genie back in the bottle.

I think everybody that’s into deep learning or knows anything about the AI space in general, knows that the rate of evolution is accelerating dramatically. There is no doubt the next 10 to 15 years — above and beyond the numb-nut that we just elected — the next 10 to 15 years are going to a challenging transformation on so many different levels.

You are absolutely right. This is a combination of technology and policy, the human experience, societal values all coming together. I look at things like the social safety net as something that we should get after right away. The way that you tax on both sides of the equation to ensure that there is viability in whatever the future of work ultimately looks like, is get inefficiency out of the system.

While that may be uncomfortable for some people, history has shown when you take inefficiency out, in theory you have the ability to reinvest that into higher levels of productivity.

Interestingly enough in your blog this morning, you made the comment that in the 30 hours of productivity that has been gained as a result of the first generation of the Internet in terms of automation and technology of the doubt, we’ve used 28 hours of that to watch TV.

Now hopefully as a society we will evolve our learning systems and other systems to teach our people the skills that are going to be relevant in 2050. I have five children, ranging from 30 to 10. I think every day about my 10-year-old and what skills is she going to have. I can see it my children they’re not going to work the way that I worked. They’re going to work their own way, with their own technology, in their own expectation.

I don’t fear AI. Robots, artificial intelligence will always serve mankind. While there are others that feel differently, we can sit and run around that tree all day long. We’re at a point in our society where the elimination of inefficiencies… You heard Tyler yesterday talk about, it took him a year to get his book out. How moronic is that in today’s society? Now his business is evolved because he’s able to leverage big data and technology to get his content out and he eliminated middlemen that made the system better.

FW: I want to talk a little bit about social safety net. Maya, when you and I talked a couple weeks ago, you told me that you thought that Social Security could be a model or framework for thinking about how to potentially protect people who may be put out of work and will have difficulty finding new work. These drivers for example, these 4 million drivers who in theory we’re going to put out of work. Can you elaborate a little bit on that thinking?

MR: Actually, I want to elaborate even more. Social insurance generally is an important asset that most Americans don’t even — it works so well that most people don’t even know how important it’s for our society. I think unemployment insurance and Social Security particularly, those two forms of social insurance will be especially important in this regard.

I just told you the population groups in the state will likely be more impacted by the loss of jobs due to autonomous vehicles. Why wouldn’t policymakers want to look at making unemployment insurance automatic for people in certain job categories, or for people in certain states, based on what we know is coming.

That would be one way where policymakers can take a leadership role. By the way we are not Luddites. We don’t think that we’re going stop to technology. In fact, we’re not even arguing for it, because there are too many peerless things in terms of the impact on our planet and other things that we need to be thinking about.

At the same time, I would urge Silicon Valley technology leaders to think strategically about the impact these technologies will have on the stability of our society and our economy, and to collaborate and to advocate as much as possible for the preservation, protection and the strengthening of our social safety net, particularly social insurance programs like unemployment insurance and Social Security.

Now, moving on, a large conversation here has been about basic income. On the other side of the country and by the way when I got to Washington DC in 1997, one of my first jobs was on the House Ways and Means Committee on the Social Security subcommittee of the House Ways and Means Committee. This is one of my areas of deep expertise.

We have the infrastructure to do a basic income now and it is called the Social Security program. It’s highly efficient. The administrative cost are less than .01 percent, which is better than you would get on the private market, and everybody has to have a Social Security card.

There is already a financing vehicle for how you actually input the funds into a trust fund that then goes back and gets paid out to workers. It would be nothing to basically design a system for basic income on top of the Social Security program.

There are already people who are arguing for strengthening the Social Security program, because they’re looking at the retirement crisis. But on the other side of the spectrum a lot of people are also looking at the social security program as a way to provide a basic income ironically enough for kids. Providing every kid born in America with what they call a baby bond or an account that can then be seeded over time and then be paid out to support them as they matriculate to get an education or a job or whatever.

FW: Maya, can I ask you a question? These concepts around unemployment insurance, social insurance, Social Security, do they limit the amount of work that you can do before you can collect — let’s take Social Security. I’m going to retire. I’m 65, I’m 67 whatever. Can I still work and collect Social Security?

MR: You can.

FW: Is there a limit to how much money I can make?

MR: No. There’s not a limit to how much money you can make but what is interesting about working and still collecting social security benefits is they will tax you on your work. They will tax that social security benefits based on the amount that you earn.

FW: The reason I ask is, you mentioned the retirement crisis and Steve one of the things that your platform does, is it allows the company to have a labor cloud of full-time workers, contractors, freelancers and also retirees that they can then access when they need it. So…I’m 65. I retire from 40 years of working at Boeing. I design the 727, 737 and 777. There’s a valve issue on one of those planes. You can get to me and I’m retired, playing golf, playing cards, and you can get me back, fixing the valve issue. Right?

SD: Well think about it. I mean you are your own enterprise. How do you manage your retirees? What do you do? What do you have in a database? How are you engaging with work? Our air traffic controllers get put out to pasture at a certain age, and they don’t get brought back in to do mentoring, which is just brain-dead. There are so many industries that look at labor costs and go, “Ooh, when I’ve gotten to a certain point I kinda want to get you out of here, because either that product that’s no longer being sold…”

Look at Sun Microsystems here in the Bay Area. There are not a lot of people buying their hardware anymore. If you’re in SPARC and Solaris and that was a big job back in the ’90s you’re put out to pasture.

The point here is, how do you take people that know their stuff, that have passion for what they do, that have intelligence, and make that connection seamless? Regardless of how you’re engaging with them. That’s what we’re focused on, and that’s what the technology enables now.

FW: I just want to talk about the retirement crisis for one more second. Is it your view, that instead of somebody retiring, you’re working, you’re not working, it’s going to be more of an amorphous change? You’re going from working full-time, nine to five, to working a little bit, but not nine to five anymore, maybe you’re not going to the office anymore.

You think that we’re just going to change that? There’s not going to be the hard stop, where the company throws the party, and you get some sort of gift for your 30 years, and then you go home and say, “Holy shit, I’m done.”

SD: You know what, if you ask 50 enterprises that are transforming, I just came from the executive leadership at Hewlett-Packard which has employed hundreds of thousands of people through the years. One of the iconic American companies. The business model was no longer competitive. It had to evolve. It had to change. For the security of the workers. For the security of the value. There was a great conversation earlier about shareholder value versus stake-holder value.

FW: Yeah, that was great wasn’t it?

SD: That was a great dynamic. I’m not equating what we faced at Hewlett-Packard as a shareholder issue. It was partially a shareholder issue, but inside the shareholder issue were all the stake-holders. I have a tough time separating the two.

FW: Let’s go to questions.

Delegate: Thank you. OK, so I’m sure you can hear from my accent, I’m from France. I’ve been hearing a lot about, all morning, basic income, social welfare, whatever. We’ve had that for 30 years, you know, everything you can dream of, we give. We still have very high unemployment rate. We have social unrest. We have the far right that has been growing to an extremely high rate. I really don’t agree with the fact that it’s the right way to approach it. I think you should you should look at what’s happening with Europe, and draw conclusions for the US.

From my perspective, I think education and training people is really solving the root of the problem. Not just throwing money at people, and hoping they would find purpose on sitting on their hands.

MR: In terms of our report, we actually argue for more education. We argue for affordable post-secondary education. Not just college, but also job trades and other things that can allow people to develop their skills, and also their critical thinking ability.

We’re also very supportive of entrepreneurs. The idea that entrepreneurship can be a vehicle for job creation, and that’s why those skills, those fundamental educational skills, are important.

At the same time, I think that it would be a critical mistake, for us not to think about how to actually make sure that the supports are there, to protect against the shock to the labor market, and to our economy in terms of the job loss. It would also be, I think, short-sighted of us to not think about a basic income.

By the way, it’s not a basic income where everybody gets the same amount. To me, that doesn’t even make sense. We’re actually talking about a basic income that’s progressive. It’s tied to the amount of wealth or income that you have in your household. It depends on things like your education level, your skill level, and your income level, and wealth level.

It’s important that we think about this holistically, and not just assume that education is going to be the panacea.

FW: Stephen, last word.

SD: Yes!! on education. I mean my God! This is absolutely it. We have to train our people for the skills that are going to give them an opportunity to make a great living. Be personally fulfilled. To follow their passions. We have to train them. The CEO of Accenture put a note out yesterday on the digital enterprise and the future of that. Accenture, which is a company which is made up of skilled professionals, made a comment that automation has taken away 17,000 jobs at Accenture and no one lost their work, because they had to consciously retrain those people into the skills that matter in the world moving forward.


Leave a Reply