The Second Machine Age: Work, Progress, and Prosperity in a Time of Brilliant Technologies (32 page)

The Peer Economy and Artificial Artificial Intelligence

Changing the subsidies and taxes on labor might seem like a short-term solution. After all, isn’t the second machine age defined by relentless automation that will lead to a largely or completely postwork economy?

We’ve argued here that in many domains it is. But, as we’ve also hopefully shown, people have skills and abilities that are not yet automated. They may become automatable at some point but this hasn’t started in any serious way thus far, which leads us to believe that it will take a while. We think we’ll have human data scientists, conference organizers, divisional managers, nurses, and busboys for some time to come.

And as we discussed previously, people still have much to offer even in heavily automated domains. Although no person now can beat the best chess computer, for example, the right mix of human and digital labor easily beats it. So it’s not the case that people cease to be valuable the instant computers surpass them in a domain. They can be enormously useful once they’ve paired up to race
with
machines, instead of against them.

We see this even in heavily automated fields like computer search. As Steve Lohr explained in a March 2013
New York Times
story,

[W]hen Mitt Romney talked of cutting government money for public broadcasting in a presidential debate last fall and mentioned Big Bird, [Twitter] messages with that phrase surged. Human judges recognized instantly that “Big Bird,” in that context and at that moment, was mainly a political comment, not a reference to “Sesame Street,” and that politics-related messages should pop up when someone searched for “Big Bird.” People can understand such references more accurately and quickly than software can, and their judgments are fed immediately into Twitter’s search algorithm. . . .

Other human helpers, known as evaluators or raters, help Google develop tweaks to its search algorithm, a powerhouse of automation, fielding 100 billion queries a month[.]
²³

So even though the algorithms are getting better, they can’t do it alone. This insight has led to new, technology-based ways to organize and accomplish work.

In the middle of the last decade, the online retail giant Amazon realized that there were more than a few duplicates among its millions of pages describing products for sale. Algorithms alone didn’t do a great job of finding them all, so a team led by employee Peter Cohen built software that showed possible duplicates to human beings and let them make the final determination.
²⁴
Cohen and Amazon soon realized that this was a generally useful innovation. It took a large problem (finding the duplicates among millions of pages), broke it down into many small tasks (are these two pages duplicates?), sent the tasks out to a large group of people, collected their responses, and used them to make progress on the problem (eliminating the duplicates).

The software was originally intended only for internal use, but in November of 2005 Amazon released it to the public under the name Mechanical Turk, in honor of a famous eighteenth-century chess-playing ‘robot’ that turned out to have a human inside it.
²⁵
The Mechanical Turk software was similar to this automaton in that it too appeared to accomplish tasks automatically, but in reality made use of human labor. It was an example of what Amazon CEO Jeff Bezos called “artificial artificial intelligence,” and another way for people to race with machines, although not one with particularly high wages.
²⁶

Mechanical Turk, which quickly became popular, was an early instance of what came to be called
crowdsourcing
, defined by communications scholar Daren Brabham as “an online, distributed problem-solving and production model.”
²⁷
This model is interesting because instead of using technology to automate a process, crowdsourcing makes it deliberately labor intensive. The labor is provided not by a preidentified group of employees, as is the case with most industrial processes, but instead by one or more people (often many more), not identified in advance, who choose to participate.

In less than a decade, crowdsourced production has become an important phenomenon. In fact, it’s given rise to a large new crop of companies, often grouped together as the ‘peer economy.’ Peer economy companies satisfy their customers’ requests by crowdsourcing them. Some of the graphs you see in this book, for example, were generated or improved by people we’d never met before. We found them by posting a request for help with the task to TaskRabbit, a company founded by software engineer Leah Busque in 2008. Busque got the idea for TaskRabbit after she ran out of dog food one night and realized that there was no quick and easy way for her to use the Internet to find (and pay) someone willing to pick some up for her.
²⁸

That same year, Joe Gebbia, Brian Chesky, and Nathan Blecharczyk also launched a website that used the Internet and the crowd to better match supply and demand. In their case, the demand was not for help with a task, but instead for a place to stay. The site, Airbedand-breakfast.com, allowed people to offer rooms in their homes to visitors; it grew out of an experience that Gebbia and Chesky had offering space in their apartment to attendees of a 2007 design conference in San Francisco, where affordable hotel rooms were scarce.

The service they built, which was renamed Airbnb.com in 2009, quickly became popular. On New Year’s Eve of 2012, for example, over 140,000 people around the world stayed in places booked via Airbnb; this is 50 percent more than could be accommodated in all the hotels on the Las Vegas Strip.
²⁹
TaskRabbit also grew quickly; by January 2013 the company was reporting “month-over-month transactional growth in the double digits.”
³⁰

TaskRabbit allows people to offer their labor to the crowd while Airbnb lets them offer an asset. The peer economy now includes many examples of both types of company. Crowdsourced labor markets exist in specific domains like programming, design, and cleaning, as well as for general task execution. And people now use websites and apps to rent out their cameras, tools, bicycles, parking spaces, dog kennels, and almost anything else they might own.

Some services bring these two models together and let people offer a combination of labor and assets over the Internet. When Andy needed to have his motorcycle towed to another state in 2010, he found the right person for the job—someone with both time and a trailer on their hands—on uShip. Lyft, founded in 2011, allows people to effectively turn their cars into taxis whenever they want, giving crosstown rides to others. In an effort to avoid opposition from taxi regulators and other authorities, Lyft does not set fees or rates. It instead suggests to customers a ‘donation’ that they should offer to the person who just gave them a lift.

As the story of Lyft highlights, there are many legal and regulatory issues that will need to be resolved as the peer economy grows. While we certainly acknowledge the need to ensure public safety, we hope that regulation in this new area will not be stifling and that the peer economy will continue to grow. We like the efficiency gains and price declines that crowdsourcing brings, but we also like the work that it brings. Participation in services like TaskRabbit and Airbnb gives people previously unavailable economic opportunities, and it also gives them something to do. It therefore has the potential to address all three of Voltaire’s “great evils,” and so should be encouraged by policy, regulation, incentives like the ETIC, and other available levers.

The peer economy is still new and still small, both relative to GDP and in absolute terms. In April 2013, for example, TaskRabbit was adding one thousand new people each month to its network of approved task completers.
³¹
This is encouraging, but that same month there were nearly 4.5 million Americans who had been out of work for at least twenty-seven weeks.
³²
Comparisons like this strongly suggest that crowdsourcing does not yet play a significant role in reducing unemployment and bringing work to the economy as a whole.

This fact does not mean that the peer economy should not be encouraged and supported. Quite the opposite. The best solutions—probably, in fact, the only real solutions—to the labor force challenges that will arise in the future will come from markets and capitalism, and from the technology-enabled creations of innovators and entrepreneurs. Peer economy companies are examples of innovations that increase the value of human labor rather than reducing it. Because we believe that work is so important, we believe that policy makers should encourage such creations.

Wild Ideas Welcomed

We’ve discussed the future and how to shape it with a variety of technologists and labor leaders, with economists and sociologists, with entrepreneurs and retail clerks, and even with science-fiction authors, and we’ve been impressed with the breadth of ideas offered. This brainstorming is valuable because we are going to need more novel and radical ideas—more ‘out-of-the-box thinking’—to deal with the consequences of technological progress. Here are a few of the ideas we’ve heard. We include them not necessarily to endorse them, but instead to spur further thinking about what kinds of interventions will be effective as machines continue to race ahead.

• Create a national mutual fund distributing the ownership of capital widely and perhaps inalienably, providing a dividend stream to all citizens and assuring the capital returns do not become highly concentrated.

• Use taxes, regulation, contests, grand challenges, or other incentives to try to direct technical change toward machines that augment human ability rather than substitute for it, toward new goods and services and away from labor savings.

• Pay people via nonprofits and other organizations to do ‘socially beneficial’ tasks, as determined by a democratic process.

• Nurture or celebrate special categories of work to be done by humans only. For instance, care for babies and young children, or perhaps the dying, might fall into this category.

• Start a ‘made by humans’ labeling movement, similar to those now in place for organic foods, or award credits for companies that employ humans, similar to the carbon offsets that can be purchased. If some consumers wanted to increase the demand for human workers, such labels or credits would let them do so.

• Provide vouchers for basic necessities like food, clothing, and housing, eliminating the extremes of poverty but letting the market manage income above that level.

• Ramp up hiring by the government via programs like the Depression-era Civilian Conservation Corps to clean up the environment, build infrastructure, and address other public goods. A variant is to increase the role of ‘workfare,’ i.e., direct payments tied to a work requirement.

Each of these ideas has promising aspects as well as flaws. We don’t doubt that there may be other ideas that would be even more effective.
*

Of course, theorizing alone has its limits. Perhaps the best advice we can give is to encourage policy experimentation and seek opportunities to systematically test ideas and learn from both successes and failures. In fact, there are individuals, industries, and even whole nations where some aspects of second-machine-age economics are visible today. There are lessons to be learned. For instance: How do lottery winners react to not having to work anymore? (Hint: not always well.) What can we learn from industries with a concentration of high-income superstars like professional sports, motion pictures, and music? What challenges and opportunities do citizens of nations like Norway and the United Arab Emirates face when they have access to enormous wealth as a birthright via sovereign wealth funds? What were the institutions and incentives that helped some children of wealthy landowners in the seventeenth century go on to lead happy, inventive, and creative lives, while others did not?

In the coming decade, we will have the good fortune to witness a wave of astonishing technologies unleashed. They will require changes in our economic institutions and intuitions. By maximizing the flexibility of our systems and mental models, we will be in the best position to identify and implement these changes. A willingness to learn from others’ ideas and adapt our own practices—to have open minds and open systems—will be the hallmarks of success.

*
The state of Alaska, however, set up a form of guaranteed income for its residents in 1980, when it passed legislation establishing universal dividends from its Permanent Fund. The Fund was set up in 1976 to manage the state’s share of its abundant oil wealth; four years later, Alaskans decided that a portion of this wealth should be distributed each year in the form of dividend checks.

*
We’re interested in hearing which ideas you like best, and others you would like to suggest. Contact us at www.SecondMachineAge.com to share your insights.

“The machine does not isolate man from the great problems of nature but plunges him more deeply into them.”

—Antoine de Saint-Exupery

I
T
’
S
ONE
OF
HUMANITY
’
S
most ancient fantasies: that someday we can all have our material needs fulfilled without drudgery, freeing us to pursue our true interests, amusements, or passions. And that someday, no one will have to toil at an unpleasant task because food, clothing, shelter, and all the other basics for living will be provided by automated servants that do all our bidding. It makes for some great stories. But for most of history, they’ve been just that: legends and myths populated by fantastical automatons made of clay (like the Jewish golem or Norse giant Mokkerkalfe, built to battle Thor), gold (in the
Iliad
, Homer describes the servants and self-driving tripods built from the precious metal by the god Hephaestus), or leather and wood (the flesh and bone of the artificial man made by craftsman Yanshi in the ancient Chinese Liezi text). The materials change, but the dream remains the same.