In a recent episode of the podcast I’ve Got Questions, Sinead Bovell makes the observation that companies currently making layoffs worry less about AI's current capabilities than about the possibility that entire workflows may be about to become obsolete. Her vision of an AI economy privileging adaptability, learning, and entrepreneurialism over traditional expertise sounds empowering - a meritocratic future where nimble generalists thrive. Yet, this framing obscures a darker trajectory: the systematic degradation of employment quality and the acceleration of inequality that such an economy practically guarantees.
Bovell correctly identifies that skills will trump experience in an AI-mediated workplace. But what does this actually mean for workers? It means perpetual re-skilling - an endless treadmill of learning that places the burden of adaptation entirely on individuals rather than institutions. When workflows become obsolete overnight, workers don't just change jobs, they lose the accumulated value of their expertise. A decade of specialised knowledge becomes worthless, and the experienced professional finds themselves competing on equal footing with the recent graduate, except the graduate carries less financial obligation and undercuts them.
This constant devaluation of experience directly parallels today's employment statistics, which tell a misleadingly optimistic story. Low unemployment figures mask the proliferation of precarious gig work contracts, zero-hours arrangements, and perma-temp positions that offer neither security nor benefits. These aren't aberrations, but structural features of an economy that treats labor as infinitely flexible and disposable. The AI economy will amplify this pattern exponentially.
Consider what "entrepreneurialism" means in this context. It's not the romantic vision of garage startups but rather the normalisation of workers functioning as atomised businesses-of-one, stripped of collective bargaining power, employment protections, and social safety nets. When everyone must be entrepreneurial just to survive, entrepreneurship ceases to be opportunity and becomes obligation - a tax levied on existence.
The quality of employment deteriorates because AI doesn't just automate tasks, it fragments work into micro-components that can be distributed, monitored, and compensated with algorithmic precision. The result is hyper-optimisation that squeezes every inefficiency from human labor while siphoning value upward to platform owners and capital holders. Workers become interchangeable modules in systems they neither control nor understand.
Bovell's analysis touches on this reality (regarding US-style employer-based health insurance) but doesn't fully reckon with its implications. If AI transformations prove as dramatic as predicted, we are not facing gradual adjustment, but potential catastrophe for working populations. The inequality already rising will become chasm-like. Those with capital to invest in AI systems will capture exponential returns. Those selling only their labour, however adaptable and entrepreneurial, will compete in ever-more-precarious conditions for a shrinking share of value.
The question isn't whether workers can learn to adapt to an AI economy, but whether that economy will provide conditions worth adapting to. An economy that demands constant reinvention while offering diminishing security and compensation isn't sustainable, it's extractive. Unless we fundamentally restructure how AI-generated productivity is distributed, the future won't be one of opportunity through adaptability but rather a broad collapse in quality of life, dressed in the language of innovation.
The real transformation required isn't in worker skills, but in economic architecture itself. Without addressing who owns AI systems, who captures their value, and what obligations exist to those whose labor they displace or devalue, we are simply describing inequality's acceleration in aspirational terms.
