Gigged Read online

  CHAPTER 6

  UBER FREEDOM

  After quitting his programming job to become a full-time freelancer with Gigster, Curtis no longer stuck to his early morning routine of taking his laptop to Starbucks. Seeking variety, he had scoped out all of the cafes near his apartment that had dependable WiFi. Sometimes he worked at the library. At others, he might work at the park or a bar. On his own schedule, he roamed between these places, and he was happy doing it. Two months into his freelancing career, Curtis was earning as much as he had in his full-time job: between $10,000 and $12,000 every month. He also now had time to hit the gym in the middle of the day, meet his girlfriend on her lunch break, and plan multiple vacations. For him, Silicon Valley’s utopic description of the gig economy seemed completely true.

  At his 9-to-5 job, Curtis had hated everything except his data-mining work: the office politics, the long chains of command, and the “selling” and self-promotion required to advance in the organization or do something new. But on Gigster, there was none of that. If he took on projects and did them well, his ranking, what the platform called his “Karma” score—based on how many projects he successfully completed—increased. And as his score got stronger, Gigster’s algorithm “trusted” him with more and more interesting projects. It was advancement without all of the extra stuff required to succeed in a traditional career.

  Almost every gig economy company had created a similar rating system. Uber asked riders to score drivers on a five-star scale after every ride (drivers did the same for riders). Handy, the gig economy cleaning company, used the same scale. Upwork allowed customers to leave workers comments and star ratings that showed on their profiles. Because gig economy companies didn’t have managers who knew workers, they relied on these ratings to algorithmically dole out rewards and punishment, such as “deactivation,” the gig economy term for getting fired (e.g., removed from the platform). Rating systems like these could replicate prejudice or feel arbitrary to some workers, but Gigster’s system worked for Curtis. Not only was he able to land enough jobs to make a living, but they were increasingly interesting. Sometimes he learned new skills in the process of completing them.

  Of course, there were some downsides to Curtis’s new gig economy lifestyle. He had to work every hour that he got paid. There were no more paid hours spent watching video game sites, and no more free snacks. When he got called for jury duty in March, he lost a week of income. Though there’s no federal law that requires employers to offer paid leave for jury duty (some state laws do), more than 60% of US workers, and 81% of US professionals and managers, collect a paycheck while they serve.1 As an independent contractor, Curtis didn’t have that luxury.

  But these downsides were manageable. Curtis had a year’s worth of savings that, because he had made money every month through Gigster, he had not even touched, and so jury duty didn’t threaten his ability to buy groceries. It meant he’d spend the next month catching up on projects. The loss of free snacks and a guaranteed paycheck were far outweighed by the freedom and challenging work that accompanied his new gig.

  By April, seven months into his freelance career, Curtis no longer wanted a job at a startup. “I don’t see what it can offer me that is better than my current situation,” he told me. “The risk [with joining a startup] I think is actually greater than freelancing, because startups pay you less, and equity is pretty much worth nothing.” I reminded him of rare cases like Uber, in which a startup becomes worth tens of billions of dollars, making early employees with equity instant millionaires. But Curtis preferred to focus on his regular paychecks rather than the odd chance of a huge payout. Freelancing suited him.

  This was the gig economy that its boosters described, and Curtis proved that it could indeed work out wonderfully.

  * * *

  The same personality traits that made Abe a competent server—a loud, faux confidence and a natural ease with strangers—also worked well for him as an Uber driver. When he was driving for Uber, he played “old-school” music, always offered gum, and sometimes offered shots of whiskey. “Pretty much whatever they wanted to do, I was up for it,” he said. “Like, they wanted to drink in the backseat, that’s fine by me. Just don’t make a mess. Whatever.” Videos Abe posted on social media showed passengers dancing in the backseat, Abe’s head bopping to the beat as he gripped the steering wheel. He called this “the Uber difference.”

  Uber doesn’t allow riders to request specific drivers, but Abe created a system with his best passengers that he called “shot-gunning a ride.” Instead of requesting a ride in the app, these clients would first call Abe directly. After entering his car, they’d open the app and request a ride. Since Abe was the closest driver in this scenario, Uber would almost always route the request to him.

  Abe mostly drove at night, and he usually started near the same row of popular clubs and bars where his friend had first shown him how to work on the Uber app. His customers were often drunk, “Which I actually prefer,” he said, “because they are easy to get along with.” Apparently the feeling was mutual: According to emails he showed me, Abe scored a 4.9 rating on Uber’s five-star system.

  Uber offered him the same deal that it had offered the friend who signed him up. If he referred a driver to the platform, he’d earn a $200 bonus after that driver’s twentieth ride (Uber made variations of this deal in different cities and during different time periods). Abe had plenty of experience getting people to sign up for things, though most of it was related to what he eventually learned were pyramid schemes. Convincing people to drive for Uber, a company that would actually pay them, couldn’t be more difficult than getting them to join GIN. In April 2015, Abe made a Facebook page called “Uber Freedom” (“because it’s Uber, and it gives you freedom”) and started posting about the thrills of being an Uber driver. His hope was that others would sign up using his referral code, which he would promote through the same Facebook page.

  His first video showed a large blond dog in the back of Abe’s Nissan Altima. “This is Waylon, my newest Uber rider,” Abe says as he films with his phone. “I’m going to give you five stars, Waylon.”

  True to the “law of attraction” that he’d learned through GIN, Abe was absolutely sure other people on Facebook would want to imitate this new Uber lifestyle he’d chosen.

  * * *

  Two years had passed since Kristy’s husband lost his job at the Nestlé factory. Though he’d gone back to school to finish his high school education, he still hadn’t been able to find a job. Kristy meanwhile had become highly proficient at making money on Mechanical Turk: In both 2011 and 2012, she earned more than $40,000 on the platform. That was before taxes, but it was still an astounding amount relative to other Mechanical Turk workers.

  According to a 2016 report from the UN International Labour Office, “crowd workers” like Kristy, 40% of whom rely on crowd work as their main source of income, on average earn between $1 and $5.50 per hour.2 The median hourly earnings of Mechanical Turk workers based in the United States is $4.65 per hour, while for Mechanical Turk workers based in India, it is $1.65 per hour. Based on a 40-hour work-week (which she often exceeded), Kristy was earning more like $20 per hour. She had achieved this relatively high rate by learning how to find the best jobs and how to set up systems that made them easier to complete.

  Beginners on Mechanical Turk were pretty much useless, because they didn’t qualify for higher-paying tasks that required a certain amount of past work to access, tended not to apply any strategy to the tasks they chose, and didn’t always understand how to complete them efficiently. I knew because I was one. Curious to understand how Mechanical Turk worked, I logged on one day and tried the work myself.

  The site has none of the polish of Amazon’s other products. It looks old, like an internet forum from the early 2000s, which perhaps suggests its low placement on Amazon’s list of priorities. Human Intelligence Tasks (HITs), as they’re called, appear in a dashboard. You can sort them to find those ava
ilable for your “qualifications,” which can include attributes such as being based in a certain country or having successfully completed a certain number of tasks. When you choose a HIT, based on a short description like “transcribe 35 seconds of media into text,” you complete it inside of the platform and earn a sum every time you hit “submit.”

  It only took me five minutes to set up an account and just a few minutes after that to locate tasks for which I qualified. But it took me almost an hour to make a dollar. Most of the tasks I selected involved taking surveys for academic studies or labeling things. The task I spent the most time on was one posted by a Microsoft researcher who was building image recognition software and needed to “teach” the algorithm how to spot and name objects. One by one, I labeled a series of hundreds of animal pictures. Each slideshow had five photos of the same type of animal in a different type of scene. Each photo in these slideshows had 11 pages of labels, which meant that one slideshow took 55 clicks to complete. I got paid $0.05 every time I completed a slideshow.

  After a few rounds, I found myself hoping for bird slideshows. Birds were usually outside and alone when photographed, which meant that most of the labels in my arsenal—bed, person, window, table, ball—did not apply. I only needed to click and drag one label: bird. A dog in a car’s passenger window, on the other hand, required the labels “car,” “mirror,” “dog,” and “person.” If that person happened to be on the phone, it also required “cell phone,” and if there was traffic on the street, it might require “motorcycle.” Birds, by requiring just one label, saved me valuable seconds, and my wrist valuable exertion, as I flipped through the slideshows. After two hours—the time it took before I got dizzy and moved on—I had completed 61 slideshows and made $1.94 per hour. From this perspective, Kristy’s $40,000 earnings seemed incredible.

  Some of her more remunerative work came from employers who posted hundreds or thousands of tasks at a time that could be completed in rapid succession. Kristy would install small software programs that allowed her to complete, say, a simple categorization task by hitting a key on her keyboard (“y” for yellow or “b” for bird) rather than clicking a mouse. Categorizing an item every five seconds for an hour at $0.03 per image would pay $10.80 per hour. She also took on more complicated tasks that paid better. Writing descriptions for product sites, for instance, could pay $1.50 per paragraph. So if she did one every five minutes, she would make $18 an hour. It was a matter of doing the work quickly and sticking with it for a long time.

  Turker Nation had a forum where workers alerted each other about these “good work” opportunities, which paid well and could be completed in large batches. To make sure that she didn’t miss any of them, Kristy set up an automated system that, when a new “good work” task was posted, would check to see how much it paid and whether she met its qualifications. If she was eligible for a task that paid $0.05, her computer would alert her with a “ping” noise. If she was eligible for a task that paid between $0.05 and $0.25, her computer would sound an alarm that sounded like a laundry machine finishing. If she was eligible for a task that paid more than $0.25, a siren would sound.

  No matter where Kristy was in her house, if she heard the alarm go off, she would run to her computer. There were thousands of other Mechanical Turk workers who were competing with each other to grab the high-paying work, which was assigned to whoever could claim it first. Kristy would sleep in her office so that she could listen for the alarm to go off at night without waking her husband up. When she spotted good tasks, often through her alarm system, she used an automated tool to keep her queue full with the maximum 25 tasks that could be assigned to her at one time, and then worked furiously to finish them and grab more before they were snatched by other people.

  One of the tasks she didn’t like to miss was answering questions from Amazon’s Q&A service. These were posted every 15 minutes, and there were two aspects that made them good tasks. The first was that people often asked the same questions, and Kristy had compiled a spreadsheet of answers that made these common questions quick to answer. She could get through a batch of several hundred in about five minutes. The second was that, to incentivize good work, each month Amazon paid a bonus of a few hundred dollars to the worker whose answers received the highest number of “thumbs up” votes from users. Each question might only pay a penny, but this bonus was significant. It meant that Kristy never wanted to miss a batch. Her routine was to listen for the alarm, complete the batch in five minutes, take ten minutes off, and then get back to work when the next batch of questions dropped.

  Another Amazon task that she prioritized was one that allowed customers to take photos of products in stores in order to find the same product on Amazon. It was how the company encouraged customers to comparison shop, but not everyone used this feature of Amazon’s app as intended. When they sent pictures of genitals, Kristy sent back a link to a book called I’m Calling the Police (the system was set up so that she could only communicate through links to Amazon products). It was worth dealing with rude inquiries like these because she had found a way to earn extra money when Amazon’s users sent photos of actual products: She sent them her affiliate marketing links and earned a percentage of purchases they made after clicking on them.

  Kristy also started proactively asking clients if they needed help designing their requests, contacting them through the Mechanical Turk site. Sometimes she collected consulting fees for her advice.

  The paradigm on both the employer and worker side of Mechanical Turk was less of a relationship between two colleagues than it was two people trying to beat a system. In one common example, companies posted the same work three different times on Mechanical Turk in order to check its accuracy. If one worker submitted a different answer than the other two who completed the task, the company assumed that worker had given a bad answer and rejected her work (which meant she wouldn’t be paid for it). To beat this system on the worker side, all a Mechanical Turk worker needed were two accounts to agree with each other. Some Turkers built automated bots to submit arbitrary (but matching) work results. The bots collected payment because they agreed, while a person who had earnestly done the work didn’t get paid. All Turkers could do when someone rigged a task this way was to tell each other to avoid it.

  Kristy didn’t feel like she could leave her apartment, or even her computer, lest she miss out on an opportunity to work on good tasks. Unlike an employee at a fast-food restaurant or a cleaning company, she didn’t get paid for any downtime, and she could earn more money by working smarter and faster. The psychology was that of a game that required her to be constantly on alert. In a way, that psychology kept her going: She’d set a goal for $100 per day, and, cent by cent, she often met it.

  * * *

  In Dumas, Arkansas, Terrence busily recruited students for the class that would teach them how to succeed in the gig economy. To advertise the new Samaschool program, he took out an ad in the Dumas Clarion, the local newspaper. It used the same language as Silicon Valley: His class would turn students into internet entrepreneurs who worked for themselves.

  Among the 30 students who passed Terrence’s initial interview were farmworkers, home care workers, and a few people who were chronically unemployed. Even a local elementary school teacher showed up. Samaschool’s plan was to teach students how to use the digital freelancing website Upwork (then two sites called oDesk and Elance), one of the biggest digital freelancing websites, to find work. Jobs in research, data entry, or customer service—all of which were plentiful on Upwork—didn’t require college degrees. All Dumas’s residents needed to obtain the work, the thinking went, was some instruction on how to promote themselves effectively and an internet connection.

  Terrence held the classes in the same community technology center where he had sprinted to his job interview. Built in 2012, the tan brick building had the look of a miniature high school. It was equipped with a pod of public computers, a workforce development office, and two large, bright
classrooms in which local universities offered courses. Samaschool’s classes took place in the same classrooms, each outfitted with long rows of white desks topped with black PCs.

  As students searched for jobs in the gig economy, they weren’t always successful. But Terrence started searching on their behalf, and slowly but surely, a handful of students began working online, in a way that they hadn’t previously imagined was possible.

  A year after Terrence got started, I visited Gary Foster, one of his most successful students. Gary lived at the time in a neat trailer nested so close to the railroad tracks that it literally shook when a train passed. The door was open, and the doorbell was broken. “Hello?” I yelled into the screen.

  “Come in!” I heard from somewhere inside.

  I found Gary in a small square room filled with Tweety Bird stuffed animals—his wife’s favorite. He was sitting at a table behind two laptop computers lined up side by side. We shook hands as he pulled on a headset. It was only after we heard a high-pitched “ding!” that I understood why he hadn’t answered the door.

  “Hello and thank you for calling Sears home warranty,” he said in a calm, confident voice. “My name is Gary, how can I help you?” He then dove into a conversation about a broken air conditioner with a man who lived in New York City, pulling up reference materials and customer service scripts on his computer as he spoke.

  Until recently, Gary had worked at a local dog food plant, but he’d been laid off when the plant was sold to a new company. He’d heard about Terrence’s class when he was at the workforce center, applying for jobs. For a few months after losing his job, Gary had worked the night shift at a Tyson plant about an hour away from his home. But he hated the long, nighttime commute, which scared him. “At any moment, anything can run in front of your vehicle and tear it up,” he said. “You break down in the road out here, and you’re stuck until someone comes to get you.