Note: to pursue fueling this resource and reflection about what kind of social, political and economic rules are implemented within algorithms that will then become the foundation layer of the so called "world without work"... (and to get how far current political parties seem not to address these stakes), here comes an interesting study to exemplify what algorithmic rules can be(come) and how they implement a "way of thinking", in this case at Uber. By extension, think of course about Amazon's automated or crowdsourced services, AirBn'B, etc.
Obviously, these algorithms are already being written right "under our noses" (think about algorithmic trading, smart cities, smart "things" & "stuff", autonomous cars and drones, etc.), certainly under the radar but not not under an "algorithmic communism" basis. Not that we know about ...
Via The Wall Street journal
-----
By Elizabeth Dwoskin
Keith Bedford for The Wall Street Journal
In defending his company against assertions that Uber drivers should be classified as employees, Uber CEO Travis Kalanick often wields the algorithm. Uber isn’t a boss, he argues. It’s a software platform that balances supply and demand to connect entrepreneurs with customers.
A new academic paper pokes holes in that argument.
Researchers Alex Rosenblat and Luke Stark at the Data and Society Research Institute and New York University point out that Uber uses software to exert similar control over workers that a human manager would. The company’s algorithm uses performance metrics, scheduling prompts, behavioral suggestions, dynamic prices, and information asymmetry “as a substitute for direct managerial power and control,” they wrote.
Uber did not immediately respond to a request for comment.
The researchers, who conducted in-depth interviews with Uber drivers and studied posts in drivers-only online forums, situate Uber and similar sharing-economy platforms in a wider conversation about the trend toward employee management and so-called on demand or predictive scheduling software. Starbucks, for instance, hasn’t replaced traditional managers, but it’s among a growing group of companies that increasingly rely on software to manage worker schedules and behavior.
Bottom line: Robots aren’t stealing your job – at least in this instance – but they’re becoming your boss. And the level of control and surveillance they exert is often far greater than human management would, the authors found.
Rather than undertaking a human-driven performance review process, Uber evaluates employees according to an automated rating system. Riders enter scores into the Uber app to rate drivers with one to five stars. Back-end software tallies the scores and sends drivers regular summaries of their performance and how they stack up to their peers.
The system, the researchers wrote, empowers Uber customers to serve as “middle managers,” essentially outsourcing management. It lets Uber “achieve an organization where the workforce behaves relatively homogeneously” without needing a manager to bark orders.
Uber’s software also exerts control over when and where drivers work, the researchers noted. The company never orders workers to drive, but its software does prod them. It alerts them when the software predicts that surge pricing is due to kick in, boosting the fare by up to four times and increasing the driver’s fee.
But drivers reported that it was difficult to tell the difference between the company’s predictions and an actual surge. They often showed up at a surge location to find the area saturated with drivers and the company no longer offering to reward them with higher payments.
Thus Uber’s software is not passive but manipulates the supply of labor and shapes the marketplace as a whole, the authors argued.
Drivers told the researchers they resisted Uber by failing to reply to company emails inquiring about their whereabouts and by posting on message boards to advise other drivers to “resist the surge.” They said they didn’t want Uber to know where they planned to be for fear the company would trick them into driving elsewhere without delivering the benefit of higher fees. Essentially, the authors said, Uber drivers resist the algorithm-boss by trying to trick him – perhaps not unlike the decisions traditional employees make about what information to share with a human boss.
Despite Uber’s depiction of drivers as entrepreneurs who control their own labor, an environment in which Uber has all the information makes it harder for drivers to make decisions that are in their interest, the authors said. Uber drivers are discouraged from turning down a fare – in some cities, drivers are prodded to pick up 90% of passengers who request a pickup – and they aren’t given fare information in advance. Drivers complained that this asymmetry resulted in sometimes losing money, since some rides are too short to be worthwhile, and they have no way to know how much they could expect to earn.
Of course, many of the practices that benefit Uber and annoy its drivers also benefit customers. And like Starbucks employees and other workers whose lives are made unpredictable by such predictive scheduling software, Uber drivers are free to quit. In that sense, they are the self-determined entrepreneurs that Uber describes. But in other ways, they clearly aren’t.