Our AI future is already here—and we are nowhere near ready

Signs of thinning predate AI. Across advanced economies, labour-force participation among prime-age men has declined over several decades. In Canada, participation among men aged 25 to 54 remains below late-1980s levels. In the United States, the decline has been more pronounced. Mental-health disability claims are rising faster than claims tied to physical injury. Time-use surveys show growing hours devoted to gaming and streaming, particularly among younger men. These patterns are often analyzed separately. They should not be. Gaming is instructive here not because it offers a settled explanation for labour-market withdrawal, but because it illustrates how alternative reward systems can compete with formal institutions for time, attention, and psychological energy. Modern video games provide clear goals, rapid feedback, visible progress, and social recognition. Status is legible. Effort is rewarded quickly. By contrast, much of modern work offers delayed or ambiguous feedback. Entry-level jobs promise “experience.” Mid-career roles promise “exposure.” Progress is often opaque and recognition sporadic. The strongest empirical evidence of labour detachment concerns men without post-secondary education and regions exposed to industrial decline. White-collar professionals remain, for now, highly attached to the labour force. The risk is prospective rather than present. If AI compresses career ladders, narrows entry-level pathways, or reduces the expected returns to professional training, patterns historically concentrated in lower-skilled segments could extend upward into occupations long assumed to be insulated. The existing literature does not support a single causal story linking gaming to labour-market disengagement. But economists such as Mark Aguiar and Erik Hurst have documented that rising time spent in immersive digital leisure is associated with weaker attachment to work among certain groups. In Japan, research on NEETs—young adults not in employment, education, or training—points to the coexistence of economic stagnation, digital immersion, and social withdrawal. The point is not that gaming “causes” disengagement. It is that lower-friction systems of psychological reward can coexist with declining participation in demanding institutions. Artificial intelligence may extend that dynamic. Unlike gaming, AI is not confined to leisure. AI-assisted therapy, coaching tools, and companion systems now provide personalized feedback and reassurance on demand. From a public-health perspective, this is a breakthrough. Access to mental-health support has long been constrained by cost and supply. But incentives matter. Historically, loneliness, boredom, insecurity, and the desire for status pushed individuals toward work, education, and relationships even when those spaces were imperfect. If AI increasingly cushions those pressures—offering validation without hierarchy or delay—the pull of real-world institutions may weaken. Labour markets are social systems as much as economic ones. If participation declines, three consequences follow. Productivity growth slows as fewer individuals invest in long arcs of skill acquisition and mastery. Public finances strain as tax bases narrow and disability and income supports expand. Institutional capacity erodes as governments, hospitals, schools, and firms struggle to recruit and retain individuals willing to tolerate complex, bureaucratic environments. Policy frameworks are poorly equipped to address this risk. Regulation focuses on safety and competition. Labour policy focuses on retraining. Mental-health policy focuses on access. Few treat motivation itself as an economic variable. Mass unemployment remains a serious risk as AI capabilities accelerate. But a subtler risk may be broader participation drift—a gradual weakening of attachment to work, education, and other demanding institutions. If recent developments represent an inflection point for white-collar automation, they may also represent an inflection point for how individuals experience effort, reward, and status. The displacement risks are becoming clearer. The motivational risks are less visible but potentially as consequential. That future will not arrive with a single shock. It will arrive through incremental changes in incentives and behaviour. And by the time it feels obvious, it may already feel normal.