Throwing Rocks at the Google Bus: How Growth Became the Enemy of Prosperity
Page 6
Programmer and avowed humanist Jaron Lanier thinks the answer is for us all to start participating in the game. Instead of freely providing social media sites and apps with data, which they in turn sell to analysts, we should demand to be cut in on the revenue. In the current system, we don’t even have access to the data or the correlations that might be of value to us. Google and Facebook invisibly vacuum it up and then make money with it. In Lanier’s vision, we would not only be able to pore through our own data, but we would get paid every time a researcher or company makes use of it.34
For Lanier, the net’s bum deal is rooted in the way it was programmed from the beginning: links go in only one direction. Some early Internet architects envisioned a system of two-way links, in which any piece of content could be traced back to its source and the creator either credited or paid. Your Web site could link to my book, but I would know when it does. This wouldn’t just help content creators; it would provide the backbone for a system through which individuals could be credited—in real money—every time their data was used to make an assessment. Or they could even be cut in on the value that was generated by the new correlation.35 Since thousands of people might have contributed data to a single finding, another mechanism for “micropayments” would allow all the fractions of pennies to pass among the various parties, hopefully adding up to something substantial and ultimately—since almost everybody’s data is valuable—the reinstatement of a middle class killed by power laws.36 If you can’t beat ’em, join ’em.
Though ingenious, Lanier’s solution could actually dehumanize things even further. If we are paid chiefly for our data, then we are all performing for the machines instead of one another. We are earning money not for the ways we create value for people but for all the passive activities that happen to be data intensive. Our only value to this digital economy comes from those aspects of ourselves that can be quantified. It may solve the problem of getting a whole bunch of activity back “on the books,” but to what end? So we can register some credits on a balance sheet? Must we accept “the books”—presumably, the double-entry ledger—as the fundamental operating system?
The problem with trying to get all human activity back on the books is that the books themselves are not neutral. They are artifacts of a very specific moment in human history—the beginning of the Renaissance—when the two-column ledger was instituted and everything came to be understood as a credit or a debit in a zero-sum game of capital management. Feeding more activity to the ledger simply cedes more of humanity and business alike to a growth-centric industrial model that was invented to thwart us to begin with.
That’s the problem with any of the many new ways we have of earning income through previously off-the-books activities. On the one hand, they create thrilling new forms of peer-to-peer commerce. eBay lets us sell our attic junk. Web site Airbnb lets us rent out our extra bedrooms to travelers. Smartphone apps Uber and Lyft let us use our vehicles to give people rides, for money. Unlike many of the other platforms we’ve looked at so far, these opportunities don’t lead to power-law distributions, because a car or home can be hired only by one person at a time. As long as you’re listed on the network and have decent reviews, you should do as well as anyone else.
From the consumer’s side, these apps are amazing. If you need a ride, you can open Uber and see a map of the area along with tiny icons for the available cars. Pick a car based on its location, the driver’s ratings, and the estimated price. The driver finds you based on your own GPS location and your profile picture. Payment happens automatically, tip included. Airbnb is equally seamless. Enter a place and date and the Web site instantly renders a map with available options clearly indicated. Roll over any location to see a photo, details, and ratings for each. Book the room, and you’ll find out where to meet your host or pick up the keys. Like Uber, it’s unparalleled for choice and convenience.
For the providers, on the other hand, these services create a new watermark for how many of one’s hours and assets should be grist for the ledger and ultimately in service of some corporation’s growth. It’s as if startups are out there writing algorithms to combat inefficiency and idleness by making sure that everything everyone owns is in use all the time. The platform collects its fee for putting user and provider, rider and driver, or guest and host together and enabling a new transaction where once there was none. Our assets are their new territory. Welcome to the sharing economy. Just as Lanier would have us share our data, these new companies would have us share our homes, cars, and anything else.
Only it’s not really sharing; it’s selling. In fact, just as there used to be an Internet that ran entirely on “shareware,” there were originally free versions of these new asset-renting platforms. Couchsurfing.com created a global community of people who both give and receive space in their homes. Airbnb, its commercial successor, pitches itself the same way but operates very differently—not only do boarders pay for lodging, but the vast majority of rentals are for entire apartments. Their ads show people sharing an extra bedroom and a place at the family table, but the statistics reveal that the vast majority (87 percent) of hosts leave their homes in order to rent them.37
Homes become amateur hotels, as the original residents try to live off the arbitrage between the rent they pay, the rent they earn, and the cost of living somewhere other than home. Even if you are having trouble finding work in the digital economy, you no longer have an excuse for being entirely off the books. Just don’t let the landlord find out what you’re doing. Likewise, the amateur taxi networks of Uber and Lyft are great ways for otherwise “underemployed” vehicle owners to make a few extra bucks. There’s no reason now to leave a worthwhile asset or hour off the books—even if the underemployed are really underpaid freelancers working a whole lot of hours already. These apps are not about sharing space in a vehicle—like driving a friend to the train station—they’re about monetizing unemployed people’s time and stuff.
Although it currently has a valuation of over $41 billion,38 Uber is no more a taxi service than Airbnb is a hotel chain. These are apps—beautiful ones but ultimately very simple ones—that make their money by encouraging people to engage in freelance versions of previously regulated industries. That’s the real arbitrage opportunity here, and that’s why local cabbies and hoteliers are up in arms. They have trained, invested, and conformed to numerous regulations to do what they do. A taxi medallion, required by law, can cost several hundred thousand dollars alone. So does a hotel license. These costs and regulations were not implemented out of spite but in order to maintain fair pricing, adequate supply, and a minimum quality of service. How can a cabbie make mandatory loan and insurance payments and compete on price against an out-of-work actor with a car, a smartphone, and a few hours to kill?
Uber, for one, well knows this. One of the company’s e-mail campaigns proclaims that Uber prices are “now cheaper than a New York City taxi”—for a limited time only. It’s as if the company is giving fair warning that its predatory pricing strategy is just a temporary measure designed to put regular yellow cabs out of business, the same way Walmart undercuts local retailers. This isn’t simply a case of technology doing something better and cheaper. Uber’s pricing power is the result not of some digital magic but of the company’s immunity from medallion fees and its $3.3 billion in venture funding.39 It has the money to set low prices and be the last man, or entity, standing.
Yet again, human professionalism and skill is undervalued in a landscape that favors technological solutions and the power of capital over anything else. The London cabbie who knows even the most obscure backstreets can’t compete against an amateur with a GPS for breadth of factual data. Under the guise of restoring a human, social, sharing element to these businesses, the crowdsharing apps actually replace skills, relationships, and local businesses with automated solutions—while a central server and the investors behind it can extract the lion’s share of the revenue.
> That’s why the final indignity will be on the Uber drivers themselves, when they are replaced with the automatic cars currently in development by Uber investor Google. The app will orchestrate the movements of robot vehicles even more seamlessly than those driven by humans, and Uber’s shareholders should do just as well—even better—in this more automated future. To them, the sharing economy is less a cultural ethos than part of a strategic transition toward more fully automated solutions. Peer-to-peer is not a means of including more people as value creators but a prelude to getting rid of them—first the skilled, fairly paid ones, and then the unskilled ones who took their places.
It’s a pivot we’ve seen before. The Netflix DVD rental Web site offered more choice and more convenience than the brick-and-mortar video rental stores it replaced while employing far fewer people. The company even figured out how to compete with the skilled clerks of the best stores by developing an algorithmic system of personalized peer-to-peer recommendations. The higher-skilled staff jobs were replaced with unskilled labor putting DVDs in mailers. When the company became a streaming service, even those unskilled jobs were eliminated.
It’s as if whenever we start down the path of trying to find an employment solution for people in a digital landscape, we end up in the same defenseless, jobless place. We can’t get paid for our cultural product unless we’re one of the Top 10 artists of the year. We can’t get good at any job skill without its being automated by someone with a free smartphone app. The more time and assets we can get on the books, the faster they are devalued or replaced by a new technology.
More than two thirds of job losses are now the direct result of having one’s function taken over by a machine. So far, these are mostly middle-class jobs, such as manufacturing, office assistance, and calculating. Commonsense advice to replaced workers was always for them to retrain and learn higher-level skills. Don’t be a secretary, be the boss! Instead of doing a job that consists of entirely repeatable tasks, which will one day be carried out by a mindless machine, choose a career path that requires human ingenuity and decision making.
But now that so many businesses are using data to make decisions, many midlevel executive positions are also being automated. It’s not just robots taking the jobs of warehouse workers; it’s big data and analytics engines taking the jobs of stock researchers, mortgage adjusters, and marketers. We can’t outrun our technologies’ race upward toward higher competencies. This leads most workers to go in the other direction, down toward less-skilled jobs or less-skilled versions of traditional jobs, but those jobs are hardly immune from subsequent waves of automation.
Some business futurists envision digital workers succeeding by working on the Internet itself.* Just as Wikipedia marshaled the talents of thousands of individual people working online to create an encyclopedia, corporations are beginning to use what are called crowdsourcing platforms to engage online workers in a multitude of freelance tasks.
Again, such opportunities are touted by their proponents as part of the digital revolution. The CEO of the CrowdFlower crowdsourcing platform, Lukas Biewald, explains that these platforms are “bringing opportunities to people who never would have had them before, and we operate in a truly egalitarian fashion, where anyone who wants to can do microtasks, no matter their gender, nationality, or socio-economic status, and can do so in a way that is entirely of their choosing and unique to them.”40
Crowdsourcing platforms, such as Amazon Mechanical Turk, pay people to perform tiny, repetitive tasks that computers just can’t handle yet. Workers log into one of the platforms from home or an Internet café and then choose from a series of tasks on offer. They might be paid three cents each time they identify the subject of a photo, transcribe a sentence from a video lecture, or list the items in a scanned receipt. These are invariably mundane tasks—the sorts of data entry that wouldn’t even exist were so many business processes not already tied to computer databases, and ones that will certainly be carried out by computers themselves sooner than later.
But for now, these tasks are the province of the click workers, a growing population of several million so far, who invisibly help computers and Web sites create the illusion of mechanical perfection. (That’s why it’s particularly fitting for Amazon to have named its service after the famous eighteenth-century magic trick in which a mannequin dressed as a Turk appeared to play chess. It was actually being controlled by a human chess player beneath the table.) The worker is not eliminated; he’s just invisible.
For employers, it’s a perfect realization of the industrial ideal: anyone can request work, do so anonymously, never meet the employee, and reject the results without ever paying. The labor force isn’t simply replaceable; it’s in constant flux, perpetually changing and responsible for its own training and care.
As digital labor scholar and activist Trebor Scholz has pointed out,41 in crowdsourcing there’s no minimum wage, no labor regulation, no governmental jurisdiction. Although 18 percent of workers on Amazon Mechanical Turks are full-time laborers, most of them make less than two dollars an hour. Amazon argues that the platform is all about choice and empowerment, that workers can “vote with their feet” against bad labor practices. But when even minimum-wage jobs aren’t available to many workers today, they are empowered to make only one choice or none at all.
The other answer—one I’ve argued myself—is for displaced workers to learn code. Anyone competent in languages such as Python, Java, or even Web coding such as HTML and CSS is currently in high demand by businesses that are still just gearing up for the digital marketplace. However, as coding becomes more commonplace, particularly in developing nations such as India, we find a lot of that work being assigned piecemeal by computerized services such as Upwork to low-paid workers in digital sweatshops. This is bound to increase. The better opportunity may be to use that code literacy to develop an app or platform oneself, but this means competing against thousands of others doing the same thing in an online marketplace ruled by the same power dynamics as the digital music business.
Besides, learning code is hard, particularly for adults who don’t remember their algebra and haven’t been raised thinking algorithmically. Learning code well enough to be a competent programmer is even harder. Although I certainly believe that any member of our highly digital society should be familiar with how these platforms work, universal code literacy won’t solve our employment crisis any more than the universal ability to read and write would result in a full-employment economy of book publishing.
It’s actually worse. A single computer program written by perhaps a dozen developers can wipe out hundreds of jobs. Digital companies employ ten times fewer people per dollar earned than traditional companies.42 Every time a company decides to relegate its computing to the cloud, it is free to release a few more IT employees. Most of the technologies we are currently developing replace or obsolesce far more employment opportunities than they create. Those that don’t—technologies that require ongoing human maintenance or participation to work—are not supported by venture capital for precisely this reason. They are considered unscalable because they require more paid human employees as the business grows.
Finally, there are jobs for those willing to assist with our transition to a more computerized society. As employment counselors like to point out, self-checkout stations may have cost you your job as a supermarket cashier, but there’s a new opening for that person who assists customers having trouble scanning their items at the kiosk, swiping their debit cards, or finding the SKU code for Swiss chard. It’s a slightly more skilled job and may even pay better than working as a regular cashier. But it’s a temporary position: soon enough, consumers will be as proficient at self-checkout as they are at getting cash from the bank machine, and the self-checkout tutor will be unnecessary. By then, digital tagging technology may have advanced to the point where people just leave stores with the items they want and get billed automatically.
It’s a conundrum that was first articulated back in the 1940s by Norbert Wiener, the inventor of cybernetics and the feedback mechanisms that turned plain old machines into responsive, decision-making robots. Wiener understood that in order for people to remain valuable in the coming technologized economy, we were going to have to figure out what we can do—if anything—better than the technologies we have created. If not, we were going to have to figure out a way to cope in a world where robots tilled the fields. His work had influence. In the 1950s, members of the Eisenhower administration began to worry about what would come after industrialism, and by 1966 the United States convened the first (and only) sessions of the National Commission on Technology, Automation, and Economic Progress. The six volumes it published were largely ignored, but they did serve as the basis for much of Daniel Bell’s highly regarded work in the 1970s about what he called the “post-industrial economy.” His main recommendation was to make our technological progress less “random” and “destructive” by matching it with upgraded political institutions.44