While mainstream discussions frame AI's impact on work as a manageable transition requiring 'reskilling' and 'adaptation,' deeper analysis reveals accelerating disruption that echoes historical warnings of technological unemployment first articulated by John Maynard Keynes. Recent assessments indicate that nearly 40% of global jobs are exposed to AI, with advanced economies facing impacts closer to 60%, where tasks in white-collar, administrative, and even creative sectors face rapid automation. Goldman Sachs has estimated that AI could affect the equivalent of 300 million full-time jobs worldwide, with particular pressure on entry-level positions. Reports from the IMF highlight that while some roles will be augmented, vulnerable occupations already show employment declines in high-AI-adoption regions, exacerbating skills gaps and gender disparities in exposure levels.

Economists who previously downplayed AI risks are shifting views, with surveys indicating plausible scenarios of millions of jobs disappearing faster than new ones emerge, especially if capability leaps continue. Dario Amodei of Anthropic has warned of potential U.S. unemployment spiking to 10-20% as AI eliminates half of entry-level white-collar roles within five years. This connects to broader patterns: productivity gains concentrate among AI developers and capital owners, widening inequality as consumer spending power erodes among displaced middle and working classes. Historical technological shifts allowed adaptation over generations; AI's exponential pace compresses this into years, outstripping policy responses like retraining programs that lag behind required mastery in emerging fields.

Polite discourse avoids the endpoint: without radical interventions such as universal basic income, wealth redistribution, or systemic economic redesign, mass technological unemployment risks fueling social unrest, political instability, and cascading failures in the social contract. The World Economic Forum's projections acknowledge net job creation in optimistic models (170 million new roles against 92 million displaced by 2030), yet these assume smooth transitions that ignore geographic, educational, and temporal mismatches. NYT reporting captures this evolving consensus among economists—that policymakers remain unprepared for structural labor market transformation. The fringe alarm of 'no jobs in five years' overstates the totality but correctly intuits the trajectory: AI is not merely automating tasks but reshaping the fundamental bargain between labor, capital, and society, potentially tipping fragile systems toward collapse if connections between disruption, inequality, and unrest remain unaddressed.