This week on my podcast, I read my @Medium column, "Revenge Of the Chickenized Reverse Centaurs," proposing a theory of the relationship between algorithms, interoperability and worker power (happy #MayDay!).
If you'd like an essay-formatted version of this thread to read or share, here's a link to it on pluralistic.net, my surveillance-free, ad-free, tracker-free blog:
Let's break it down. Start with "#chickenization": this is a labor economics term referring to industries that follow the model of the American poultry industry. 3/
A cartel of poultry processors have divided the country into exclusive territories, so that chicken farmers only have one market to sell their birds to. 4/
These monopolists style the farmers who supply them as "independent contractors," guarantee them no pay, but exercise control over them in ways that put the most micromanaging dickhead boss to shame. 5/
Big Chicken tells farmers which chicks to buy, what kind of coops to raise them in, when the lights go on and off, which vets they're allowed to use and which medicines the vets are allowed to administer. 6/
They even tell them who they're allowed to hire to fix their coops (specifically, they bar farmers from hiring ex-farmers who speak out against the industry).
The processors tell the farmers everything...except how much they'll be paid for their birds. 7/
This is decided only after the farmers bring them to market, and the sum is titrated to pay them enough to service their debts and raise another batch of chickens, but not one penny more. 8/
This worker misclassification and control - governed like an employee, paid like a contractor - has spilled out beyond the poultry industry. 9/
Uber drivers are heavily chickenized, with their pay calculated to let them service their car loans, insurance payments and fuel bills - but not enough to save up and quit the industry. 10/
That's chickenization. What about "reverse-centaurs"?
In AI circles, a "centaur" is a human/AI collaboration, like when chess masters and chess programs form a team that can trounce the best people *and* the best programs. 11/
In these centaur ops, the human is the "head" and the AI is the "body" - a "decision-support system," that augments the human.
Think of an Amazon delivery driver, whose work is observed and analyzed by a constellation of cameras and algorithms. These workers are the body, not the head - the AI gives the orders and the human is the dumb meat, augmenting the machine. 13/
Now we're ready to put it all together. A chickenized reverse-centaur is a worker who is misclassified as a contractor, micromanaged like an employee, and given no guarantees of pay or hours. 14/
This is end-stage app work capitalism, as with Doordash and Uber, where workers don't get to see the full amount on offer until they take the job. This lets these unprofitable companies continue to grow by offering subsidized services to customers. 15/
The app work monopolists have always relied on subsidies to grow, and this tactic switches the costs of subsidies from their shareholders to their workers. 16/
App-capitalism simps will say, "if you don't like the working conditions, just work for someone else," pretending ignorance of hundreds of years of labor activism and the principles of solidarity and collective bargaining. 17/
Luckily, workers haven't forgotten these principles and we're living in a renaissance of union organizing, in which networked communications play a central role, allowing organizers and workers to coordinate nationwide and globally. 18/
Some of the tactics that emerge from these networked labor forums are really creative. Doordash drivers created #DECLINENOW, a campaign to collectively refuse Doordash jobs forcing the algorithm to offer a higher payout for each run.
But joining a Discord server or Telegram channel for fellow Uber drivers or Starbucks baristas is just table-stakes, even if it produces a great tactic like #DECLINENOW. The next generation of networked labor organizing comes from counter-apps. 20/
Take @GigSmarter, whose Doordash driver app cracks open the job offers received by the official app to reveal the true amount the driver will be paid - something that was otherwise hidden.
The company's goal is to produce an app that lets workers survey *all* the offers being made by all the gig work companies and pick the highest-paid ones from moment to moment. 22/
That turns today's auction for who will *sell* labor most cheaply into an auction for who will *pay* the most for labor.
That's exciting, but even cooler are the "counter-algorithmic" apps. 23/
Remember when drivers started hanging their burner phones from the trees outside Amazon warehouses? They needed to make quota, but that was impossible without being assigned deliveries, and Amazon's app only assigned deliveries to drivers who were near the warehouse. 24/
Hanging a phone from a tree next to the warehouse was a lo-fi hack against the system.
The high-tech version of this comes from #Indonesia, where #tuyul apps - created by and for gig drivers - directly modify the way the companies' apps work. 26/
For example, one tuyul feature allows drivers to spoof GPS data, which means they can book fares from commuters arriving at train stations without having to jostle in (and exacerbate) the traffic jams that surround them.
Tuyul apps are incredible, embedded in wider social systems of worker solidarity, including co-ops that manage worker clubhouse/breakrooms. The best writer on this is @qadrida:
Counter-algorithmic work is happening all over. Take @trackingexposed, whose browser plugins crowdsource data that reveals the inner workings of recommendation algorithms on Youtube, Facebook, Amazon, Tiktok and Pornhub.
This is critical to understanding the way that Big Tech recommender systems shape public opinion (the group has published papers detailing the role of Facebook's algorithm in Dutch elections, and how Tiktok's algorithm promotes war propaganda in Russia). 30/
But it's also critical to labor causes, because the workers who create the videos and other creative output that drive profits for Big Tech have no way of knowing whether their work will be shown to their subscribers or blackholed by the algorithm. 31/
Hence #Algospeak, a euphemistic social media dialect that substitutes words believed to be disfavored by the algorithm ("suicide") with creative (and tortuous) equivalents ("become unalive"):
As @taylorlornez wrote in her excellent @washingtonpost story on algospeak, creative workers have built sprawling lexicons of words and phrases believed to be disfavored by the systems that they rely on to pay the bills.
These dictionaries are build on intuition and folk-belief, not science. It would be great if the companies would simply tell these workers what they should avoid if they want to get paid for their labor. 34/
We might disagree with those choices, but at least we could talk about it. That's one of @officialtoca's key demands: "Transparent and Responsive moderation":
But recommendation algorithms are the last place where the discredited idea of #SecurityThroughObscurity is taken seriously. 36/
People who should know better give tech giants a pass when they say they've invented a security system that only works if we're not allowed to know about it. 37/
That leaves us to reverse-engineer these algorithms in order to learn when creative workers' wages will be paid, and when they will be stolen.
Tracking Exposed has started to move beyond analysis and into actual seizure of the means of recommendation. 38/
Their Youchoose browser plugin lets audiences bypass the Youtube recommender and get suggestions of videos from multiple video services, whcih users can understand and reconfigure:
Counter-algorithmic work isn't just for warehouse workers, Doordash drivers, and Tiktok creators. 40/
The lockdown created a boom in #bossware, tools that turn "work from home" into "live at work," allowing your boss to automatically monitor your communications, keystrokes and eye-movements:
Without labor organizing - including counter-apps and counter-algorithms - we're *all* destined to become chickenized reverse centaurs. When it comes to seizing the means of computation, all workers stand to benefit. 42/
Indie bookstores can change your life. In 2007, I wandered into NYC's St Marks' Bookstore (RIP) and picked up a book from the recommended table: "How to Cheat at Everything," by Simon Lovell (RIP), and I learned how to spot a scam.
If you'd like an essay-formatted version of this thread to read or share, here's a link to it on pluralistic.net, my surveillance-free, ad-free, tracker-free blog: