Economists and AI pioneers are discussing the potential for an "Engels' pause" in the modern AI-driven economy. This term, coined by Oxford economist Robert Allen, refers to a period during the 19th-century Industrial Revolution in Britain where industrial output surged, yet ordinary living standards and wages stagnated for decades before broad-based prosperity emerged. AI pioneer and Nobel Laureate Geoffrey Hinton has suggested that AI could make a few people wealthy while leaving the majority poorer, hinting at such a pause.
Empirical signs of a modern Engels' pause could include productivity gains coupled with stagnant wages, as seen in some call centers where generative AI copilots boost efficiency by 30%-50% without significant wage increases. Another marker is the rising cost of complementary AI necessities like cloud computing, retraining, data access, and cybersecurity, which increase the "price of staying relevant" for workers. Job displacement and task transformation, with AI making inroads into fields like medicine, education, and finance, also serve as early indicators.
AI shares characteristics with other general-purpose technologies, like steam power or electricity, which historically brought significant growth alongside initial dislocation. Before benefits become widely shared, complementary innovations, institutional adjustments, and new skills are often required. PwC estimates AI could add $15.7 trillion to global GDP by 2030, but these benefits might be concentrated in the U.S., China, and a few foundational AI firms, potentially deepening global inequality, as the IMF estimates 40% of jobs worldwide are exposed to AI.
To mitigate an AI-induced Engels' pause, policymakers are considering several strategies. These include robust skills transition programs, similar to Singapore's SkillsFuture initiative, and treating AI infrastructure as a public good to ensure equitable access to compute and data. Additionally, discussions involve wealth redistribution mechanisms like robot taxes or Universal Basic Income (UBI) to channel AI gains towards public welfare. While the historical parallel is sobering, some argue that today's stronger social safety nets and rapid technology diffusion might shorten an AI-driven pause, particularly if policy aligns innovation with equitable deployment, for instance by leveraging AI's potential to lower costs in healthcare and education.
Ultimately, lessons from previous industrial eras suggest that without proactive governance and reforms addressing labor rights, public education, and welfare, an Engels' pause can persist, leading to increased inequality and social unrest. The outcome of AI's economic impact—whether it leads to sustained broad-based prosperity or a prolonged period of stagnant living standards for many—depends on deliberate policy choices and political will.
