When AI Becomes the Boss: The Struggle to Regulate Winner-Take-All Economics
Dianne See Morrison, (written for University of Oxford Executive Education)
What happens when AI-driven “winner-take-all economics” spreads across more industries, concentrating more power into the hands of a few AI “superpowers?”
According to AI scientist and venture capitalist Kai Fu Lee, author of AI Superpowers: China, Silicon Valley, and the New World Order, if left unchecked, AI will create a radically divided labour market. It won’t just squeeze out physical labourers whose jobs are automated, but it will also displace the middle class. The decision-making skills once thought to insulate white collar knowledge workers from job losses are looking increasingly vulnerable to AI. What will remain is an unequal job market, sharply split between a sliver of lucrative jobs for top performers, and low-paying gig jobs for everyone else in industries hardest to automate, such as home healthcare. Corporate profits for AI winners will “explode,” further exacerbating inequality by rewarding only the “elite executives and engineers lucky enough to get in on the action.” Imagine, Lee posits: how profitable Uber would be without human drivers, Apple without factory workers, or Walmart without cashiers, truck drivers or warehouse workers?
Authors Marco Iansiti and Karim R. Lakhani reach a similar, albeit more reserved, conclusion in their book Competing in the Age of AI: 'The pattern toward concentration creates increased inequality, not only across workers but also across firms, which further segments wealth, power, and relevance across markets, industries, and geographies.' This, they warn, will 'naturally' lead to a general sense of inequity, frustration, and anger, particularly in vulnerable regions and sectors.
The Gig Economy and Worker Vulnerability
Indeed, we needn’t step too far into the future to see the impact that platforms and AI are having on work, and on the implications that they raise for society and policy makers. Ever since the first Uber headed out onto San Francisco’s streets in 2010, introducing the world to a business model that would become known as the gig economy, we have been in the process of shifting tens of millions of workers to a new reality of employment. In exchange for being their own free agents, controlling when and whom they work for, workers forgo traditional employee protections and benefits—or so the original deal of the “sharing economy” went.
As the sharing economy morphed into the gig economy and platform companies responded to competition, the VC-subsidized high pay of the early days has given way to falling, fluctuating wages. Today, algorithms act as virtual bosses, tracking workers’ productivity and either penalizing them for their transgressions against efficiency or ferreting out their productivity gains to readjust KPIs. Herein lies the main point of contention between gig worker and platform: paradoxically, the flexibility promised by an AI-driven platform is also the source of a worker’s insecurity. For example, while food delivery riders in New York City initially embraced the freedom to work on their own schedules, they soon discovered that algorithms quickly recalibrated their performance expectations. As riders switched to electric bicycles to speed up deliveries, the platform adjusted delivery times and pay rates accordingly. Now, workers aren’t just expected to make faster deliveries—they are effectively forced to buy an electric bike to compete with other riders. What was sold as flexibility has quickly turned into a relentless race to meet shifting metrics, shoulder the costs of staying competitive, and surrender more and more control over their livelihoods.
Tweaks to black box algorithms to boost slowing revenues have long maddened gig workers, who are banding together and fighting back. Lawsuits against platforms, including Uber, DoorDash, Grubhub, Instacart, and Lyft, are now common. Regulators around the world are also increasingly stepping in to try to clarify whether gig workers should be classified as employees or contractors. But, as California’s experience with Assembly Bill 5 has shown, the process of deciding the status of gig workers can take years of contentious court battles to decide. Meanwhile, the “bifurcated job market” Lee predicts is already here. In 2020, DoorDash CEO Tony Xu was awarded $414 million in compensation, while some “Dashers” eked out as little as $1.45 an hour. Jeffrey Fang, a 40-year old San Francisco father made global headlines after his car—containing his two sleeping children—was carjacked during a “dash." Fang was unable to afford childcare.
The Challenge for Policymakers
Worker protections and job losses are just part of the story. Policymakers will need to rethink the social safety net—already strained in many countries as populations decline—as more workers join the ranks of the self-employed or unemployed. How should the safety net be funded if platform businesses, unlike traditional companies, are exempt from employee taxes? Should robots that replace human workers be taxed, as Bill Gates has suggested? Do platform companies enjoy an unfair advantage over traditional firms that must pay for worker protections and benefits? Is universal basic income the answer? These are just a few of the thorny questions facing policymakers.
For regulators, the work to create policy that both discourages and checks market monopolies, while still nurturing innovation, is an increasingly fraught task. AI promises a frictionless, efficient world where machines replace slow, fallible, fickle human beings. But as opaque algorithms take over decision-making and generative AI replaces workers, regulators must grapple with not only the immediate effects of these technologies but also the deeper, long-term societal consequences that they will inevitably engender. For years, Silicon Valley’s ethos—exemplified by Mark Zuckerberg’s mantra, “Move fast and break things”—has long lionised rapid growth and the pursuit of innovation over deliberative development. The challenge now is not just to ensure that, in the rush to innovate, society isn’t left behind, but to prevent unchecked innovation from irretrievably breaking the very systems it aims to improve.
(Originally written for university executive education course).