Blockchain Chicken Farm

by Xiaowei Wang

  Skynet, the unironically named government surveillance system, is also featured in the showroom. The ambitious program hopes to catch China up to the United States in the number of surveillance cameras per person. Under the umbrella of Skynet is the XueLiang or Sharp Eyes program, which implements surveillance in less developed, rural, and autonomous areas like Xinjiang. Although Skynet has nationalist ambitions for a modern surveillance state, the 2018 trade war between the United States and China has highlighted the global nature of the surveillance industry. Cameras used for Skynet are an assembly of parts, with some components coming from as far away as the Netherlands, and chips from the United States. The cameras are assembled in China, some with Chinese firmware.

  Megvii/Face++ has investors that span the globe, from South Korea to Russia to Abu Dhabi. Its technical leads have diverse, international educations. When I visit their offices, some of the people I talk to casually mention that they used to live in the United States or worked at The Wall Street Journal.

  It would have been easy to believe the technological arms race between China and the United States is real, easy to believe that the company behind China’s Skynet had an air of Soviet-era secrecy. It would have been easy because at least then a person, a company, a country could serve as the symbol of sinister surveillance. Instead, I was met with a total indifferent openness combined with the dry, surgical threat of a nondisclosure agreement. It didn’t just remind me of Silicon Valley; it was Silicon Valley.

  And in Silicon Valley, money and investment are reliquaries of devotion that allow people to transcend rules. For all the attention the Chinese Great Firewall receives in blocking Twitter and YouTube, Alibaba’s Aliyun offers a way to bypass the Great Firewall, for a fee. Other well-known surveillance companies such as Hikvision and SenseTime have a slew of foreign investment. SenseTime’s investors include Qualcomm, Fidelity International, Silver Lake Partners (based in Menlo Park, California), and Japan’s SoftBank Vision Fund. SoftBank’s Vision Fund has ties to Saudi wealth, and spans the globe in its international influence—investing in companies from Alibaba to WeWork and Slack.

  In one corner of the Megvii showroom is a display about Meitu, the beauty and cosmetics app with which you can quickly edit your selfies. The commercial is cheerful: a woman gives herself bunny ears, some blush, and lipstick. She makes herself a little paler, her eyes a little bigger, as is the fashion in East Asia—“Caucasian” features are seen as beautiful. These beauty standards, accelerated by convenient beauty app face filters, are evident in the plastic surgery ads that have now overtaken Chinese cities, which feature brown-haired, dewy-skinned white women with the ever-coveted nose bridge.

  The commercial runs, with sound effects, on repeat, every minute. Standing in front of the screen, abetted by the occasional coo and glitter effects behind her, the Megvii spokeswoman I talk to makes it very clear: Megvii doesn’t store any data, it just makes the algorithm. It is innocent, she says. What governments and companies do with it is up to them. The engineers show up every day and just do their job.

  The Megvii algorithms break down bodies and life into numbers, measurements, and parts. This kind of thinking is not new—many of us have been locked into it for hundreds of years, while grasping at an elusive, atomic sense of identity. Looking at the engineers at their desks, it can be easy to judge their ethics, to question why they continue to show up every day when Skynet videos play on loop next door. Yet, like most desk-based jobs these days, the ethical boundary becomes defined by awareness. When you have been made accustomed to solving problems by breaking them down into parts, how could you see the larger picture to know whether you’re doing harm? The world is certainly complex, but doesn’t it feel good helping law enforcement make the world safer? Why shouldn’t you trust that your work is being used by policy makers who know what they are doing?

  3.

  In a dark, dingy hallway, I see a flash of bright yellow. My Meituan driver has made it to the twenty-second floor. He has two minutes to drop off my order, and that includes the time it takes to get from the first floor all the way up here. He’s running now—past the neighbor’s electric scooter leaned against the wall—and does a soaring jeté over the wooden couch in the middle of the hallway, landing with both feet on the floor like a triumphant gymnast. He hands me my bag of cold medicine and wordlessly scurries off.

  Meituan is one of the apps that the young analyst in Palo Alto opens multiple times a day. It’s an enormously popular delivery app in China, and heralded as one of the latest innovations in China speed, a frenetic pace made possible by complex logistical coordination, the brute force of cheap labor, and the availability of low-cost mobile phones. Gone are the days of riffling through vegetables at the market, batting away old women at the store in a rush for the best produce. Everything from medicine to cooked meals and vegetables can now be delivered to you via Meituan, or scheduled to arrive at your door when you get home from work. Chinese cities are now crowded with an army of yellow-uniformed delivery drivers, riding electric scooters that weave through dense traffic and bike lanes.

  As a platform, Meituan’s business model is robust. Kai-Fu Lee, founder of Sinovation Ventures, explained what made Meituan so successful at a 2017 talk at the Asia Society in San Francisco. The delivery algorithm is aided by artificial intelligence, specifically machine learning, in crunching the massive amount of location-based geographical data and historical traffic data in order to give the best predictions and driver instructions. Because of China’s “gladiator”-like tech scene, competition is tough—Meituan fought hard against Alibaba’s Ele.me delivery service. In the end, Meituan prevailed, keeping the cost of each delivery to 70 cents, and placing swap stations for its delivery fleet’s electric moped batteries across the city.

  A few days after I recover from my cold, I find myself at a Meituan office in the county of Xifeng. It’s a funny little town that is still being formed, with buildings piled up out of the grit, clay, and gravel of its lumpy mountain geography. Even the food here has an aftertaste of dust. The town is so small that it’s not even big enough yet to be ranked on the city classification scheme.

  The Meituan office is located across the sky bridge from a sad-looking piano shop. Inside the office, two electric scooters sit, connected to power strips with tangled wires that lead to batteries and phones. One poster hangs next to a mirror, showcasing the three required styles of dress for Meituan employees: a spring/fall uniform, a short-sleeve summer outfit, and a winter jacket with coat and gloves. Everything is bright yellow. A large poster shows the best couriers for the month in Xifeng, and their awards. First place is Luo Re Ding, who gets an RMB 300 bonus (US$40, the cost of one blockchain chicken); in second place is Wang Yun, with an RMB 200 bonus. Third place gets a battery pack for their cell phone, and fourth place a waterproof cell phone case.

  It smells like stale cigarettes, sweaty socks, and take-out food. Plastic cups are strewn everywhere. Four computers sit, each with a mouse pad that shows the happy Meituan mascot—a plump kangaroo. In the corner is an enormous stuffed Meituan kangaroo with worn-out yellow fabric, on top of a pile of branded helmets. Two very nice women, looking bored on their cell phones, run the office. A few heat lamps blast hot dry air into the room, amplifying the smell of stale cigarettes. Welcome to the Xifeng branch of a company that was valued at US$55 billion for its IPO.

  I ask the women what exactly they do. They tell me they perform office tasks, dealing with admin duties and customer complaints. There are only about twenty couriers today in the city of Xifeng. Meituan is just getting started here.

  On the wall, recruitment flyers are tacked everywhere. The salary breakdown is also posted on the wall. Unlike most gig economy drivers in the United States, Meituan couriers are paid a base salary of RMB 2,000 (US$280) a month. They don’t receive any benefits besides accident insurance, so they rely on their government-provided health insurance. Each delivery earns RMB 4 on top of their base salary.

&nbs
p; Everything else listed on the salary breakdown sheet is a fine. Being late for a delivery incurs an RMB 5 fine. A bad review costs the driver an RMB 20 fine. Refusing to take an order after it is assigned in the system results in an RMB 100 fine, and being too early for a delivery is a whopping RMB 500 fine. Although the base salary initially seems appealing, the long list of fines creates a precarious existence for the workers. It parallels gig work in the United States—DoorDash drivers and TaskRabbits. In both cases, the larger system, the platform, off-loaded any operating risk to the workers, allowing the platform to shirk responsibility to customers. Through fee structures of bonuses, fines, and competition to be courier of the month, work is gamified, just like in the United States.

  When I ask the two women how long the structure has been like this, one of them shrugs. “I mean, it’s been like this ever since we started. But I don’t make the rules. I just enforce them. I’m just doing my job.”

  4.

  In 2012, a close friend and I traveled to Baotou in Inner Mongolia, a rare earth processing town. North of the town were the Bayan Obo mines, estimated to have 70 percent of the planet’s (known) rare earth deposits. Iron and fluorite surged through the earth. After being excavated, it was processed and shipped out, transported to Baotou, to be turned into cell phones, computers, and batteries.

  Inner Mongolia was my friend’s home province. The trip was his attempt for us to see the grasslands of the region—instead, it ended up being a tiring journey on buses and trains, a requiem for the last parts of nomadic Mongolian life. We drove from a copper mine to a coal mine outside of Hulunbuir, watching straggly sheep herds in the waning grassland. He’s ethnically Mongolian, but his family ended up on the Chinese side of the border during the split between independent Mongolia and Inner Mongolia as a territory of China. As an ethnic minority in China, he has a fraught relationship to the government. Early assimilationist policies, alongside forced resettlement of nomadic Mongolian herders into Han Chinese–built cities, has led to an erasure of Mongolian language and culture in the region. Mining, an industry exalted by the national government, is desecrating pastureland—these tensions were crystallized in a series of 2011 protests after a Han Chinese miner ran over a Mongolian herder.

  Carving away the hours while waiting for the next bus out of Baotou, in feverish boredom, we shared stories from the past. My friend flatly revealed that he had, at the age of nineteen, spent several years in prison after he and some childhood friends had a messy entanglement with a local bank. The incident was driven by his naive teenage wish to finally have the fiscal means to leave the dusty backwaters of Inner Mongolia, to escape the empty life that awaited him. The disappearing wild horses and the summer sounds of frogs, the coal mines being excavated, devoid of meaning. Moved by a teenage claustrophobia that I recognized in myself, he and his friends tried to get rich quick.

  He recalled the hard labor, the time spent sitting doing nothing in jail. The occasional taste of a cigarette. The paralysis he felt after he left prison, spending two years watching shanzhai Godard movies at his parents’ house, in a deep depression. Dreaming of a freedom, a kind of global, cosmopolitan world, dreaming of moving to the United States. Decades later, his life would look remarkably different in Beijing. Yet he still held his criminal record, which authorities saw as adding to his already suspect status as an ethnic minority, a migrant in Beijing. Surveillance would follow him everywhere, someone always keeping track.

  In the dusty, obscured orange sun of Baotou’s afternoon light, I felt a pang of intimacy. In a socioeconomic system that required deep chasms, my friend said it was unthinkably funny and bizarre to him that the two of us should have ever met. Him, a migrant to Beijing who had barely finished high school, an ex-convict, an ethnic minority in China. Me, a Han Chinese American expat, Harvard educated, a dutiful American citizen and taxpayer, purchaser of cell phones that aided in the creation of dusty backwaters like Baotou, the unspoken reliance on my U.S. passport and the government behind it. It is easy to feel guilty about privilege when it’s in the abstract. Instead, I looked at him and took his hand.

  In the moment that I should have felt fear toward him, a suspicion that he wanted something from me, I instead felt a strong flash of sadness. He had already recognized how the world might see him from the outside. He lived through the lens of another.

  I think of him and his lifelong awareness of being watched. We have a sense of how entangled we are in a culture of surveillance and, especially these days, how that culture proliferates with smart devices like Alexa in our homes, or as we spew clever quips on social media. The awareness of surveillance capitalism grows.

  Yet as a tactic of policing, surveillance has always been crucial in making criminality throughout history, drawing a line between those on the so-called right and wrong sides of society. And this line drawing is enabled by distilling life into arbitrary parts: class, race, gender, with the line of criminality itself constantly shifting throughout time, serving political-economic crises. “Crime went up; crime came down; we cracked down,” writes the scholar Ruth Gilmore.2

  “I don’t mind being surveilled by Alexa because I have nothing to hide” is a refrain I often hear. And while some of us might feel indignant about corporate platforms surveilling us, part of our indignation also arises because it wasn’t part of the bargain—people on the “right” side of society weren’t supposed to be surveilled. Privacy is seen as what we give up for safety, and safety is the freedom from fear.

  And in San Francisco, whenever I hear people talk about corporate, profit-driven platforms that track and monetize our data, I nod vigorously. Yet I long for a leap, for us to reconsider how surveillance has been made into a moral imperative in policing, how somehow the police state has been naturalized. Surveillance has never just been about crime, as historians and scholars like Simone Browne have shown, and it is in fact deeply tied to race, ethnicity, and white supremacist constructions of criminality. Just as platforms off-load risk onto gig economy workers, unchecked capitalism creates economic inequality and off-loads the risks and fears onto us all.

  In order for us to challenge surveillance, we will have to move beyond corporate, profit-driven platforms that track us and monetize our data, but more importantly we will have to combat our own fears and illusions of safety. We must question the culture of surveillance and carceral punishment that condition us to think living with fear is the only way of understanding we are alive. We must rethink what safety means, and what it means to build communities that allow everyone to live an unbounded life, instead of punishing people for being poor.

  As the activist Tawana Petty puts it, it’s recognizing the difference between “safety” and “security.” This work is deeply tied to transformative justice and the work of prison abolition. Until we do this work, we will not be able to move past surveillance as normalized activity and we will not be able to adequately advocate for the right to privacy for all.

  I think of my friend and the tired look on his face, the slight look of shame as he disclosed to me his past. We had known each other for a few years, traveled for countless hours. I think of the relief he said he felt after telling me, and him imagining what I would think for our entire time together. His eventual disillusionment with the United States after visiting American cities, where he felt that Americans were not as free as they imagined, but instead governed by fear in their everyday actions.

  I don’t want to live in a world where privacy is declared a human right only for a category of humans like me, and not for others. And this right to privacy is not an individualistic one of secrets and stories, but a social one that requires us to lead with trust in our daily lives. In doing so, we might even end up with a freedom from fear, the freedom we are looking for in notions of “safety.”

  There is another side to data, illuminated once we understand constructions of fear in our day-to-day lives. “Can data ever know who we really are?” asks the policy researcher and activist Zara
Rahman.3 For my friend and many others, the data on a past crime remains committed to his record. And while his life changes—he becomes a friend, a husband, a doting father, an artist, an uncle—all the data points about his past remain static. Data cannot truly represent the full spectrum of life. It remains a thin slice of the world. There is always some kind of bias built in. Yet we imagine numbers to mean something, and this creates a common tendency that the statistician Philip B. Stark calls “quantifauxcation”: the attempt to assign numbers or quantify phenomenon, as if quantitative data can offer certainty.4 Some strategies for quantifauxcation, says Stark, include saying things people want to believe, and adding opaque complexity to models, since complexity has become conflated with accuracy.

  The mixture of crime data with prediction is the realm of quantifauxcation. Xiaoli admits he can’t predict the future, but remains convinced that collecting data about past crimes in the urban village will highlight “problem areas,” helping to focus an overloaded police force. Paying extra attention to problem areas raises an observation bias in itself. The more patrols are assigned to a certain area, the more crimes are observed. And really, the numbers don’t mean much in themselves, except as markers of both how policing has replaced social welfare services and the corporate-style expectations of efficiency that are put on police officers. Xiaoli’s predictive policing reflects the circular logic that has become embedded across many cultures, enabled by technological solutionism. As the scholar-activist Ruha Benjamin puts it, “Crime prediction is better understood as crime production.”5