AI-driven productivity gains fuel job polarization and wage stagnation in exposed industries, exacerbating inequality and precarity

Indigenous epistemologies view technology as a relational tool within ecological and communal systems, not an autonomous force driving growth. The framing of AI as a neutral productivity enhancer ignores how Indigenous knowledge systems—such as the Andean concept of 'buen vivir'—prioritize harmony over extraction. These perspectives critique the commodification of labor and data, which AI accelerates by treating both as inputs for capital accumulation. Without centering Indigenous sovereignty, AI-driven economies risk replicating colonial patterns of dispossession.

Technological disruptions have historically concentrated wealth while displacing labor, from the Enclosure Acts to the Industrial Revolution’s deskilling of artisans. The current AI wave mirrors the 19th-century Luddite era, where mechanization enriched factory owners while immiserating weavers—yet mainstream narratives still frame such transitions as inevitable progress. The post-WWII Fordist compromise temporarily bridged this gap through unionization and social safety nets, but neoliberal deregulation has eroded these protections. AI’s acceleration of this trend risks repeating the failures of past 'creative destruction' without learning from them.

Non-Western economies often deploy technology within frameworks that resist hyper-individualism, such as India’s 'jajmani' system of reciprocal labor or Japan’s 'wa' (harmony) in workplace automation. In Rwanda, the government’s AI for social good initiatives prioritize equitable access, contrasting with Silicon Valley’s extractive models. These examples reveal that AI’s impact is not predetermined but shaped by governance structures—whether communal, state-led, or corporate. The Western narrative’s focus on productivity obscures these alternatives, framing them as 'inefficient' rather than as viable models for inclusive development.

Empirical studies show AI’s productivity gains are highly concentrated in a few firms (e.g., 5% of companies account for 90% of AI-driven productivity increases), exacerbating market concentration. Research from the IMF and OECD links AI exposure to wage polarization, with high-skilled workers seeing gains while low-skilled roles face displacement. The 'productivity paradox'—where firms invest heavily in AI without measurable output gains—suggests implementation challenges and misaligned incentives. Meanwhile, the 'black box' nature of AI models limits accountability, making it difficult to assess their true economic and social costs.

Artists and spiritual traditions have long critiqued the dehumanizing effects of mechanization, from William Blake’s 'dark Satanic Mills' to Buddhist teachings on attachment to outcomes. The AI-driven gig economy’s algorithmic management echoes the 'iron cage' of bureaucracy described by Max Weber, stripping work of meaning and autonomy. Indigenous art forms, such as the Navajo sand paintings, embody a cyclical view of technology as a tool for balance rather than domination. These perspectives invite a reimagining of work as a site of creativity and connection, not just productivity.

Scenario modelling by the ILO and World Bank projects AI could displace 85 million jobs globally by 2025 while creating 97 million new roles—but these are skewed toward high-skill, high-wage sectors. Futures where AI entrenches inequality risk social unrest, as seen in historical precedents like the 1930s labor uprisings. Conversely, democratic AI governance—such as participatory design or worker-owned platforms—could redistribute gains more equitably. The key variable is policy: without proactive intervention, AI will likely deepen precarity, but with it, new economic models (e.g., universal basic assets) could emerge.

Women, racial minorities, and gig workers—who are overrepresented in AI-exposed industries—face the highest displacement risks while having the least access to reskilling programs. Black and Latino workers in the U.S. are 1.5x more likely to be in jobs highly exposed to AI, yet underrepresented in tech policy debates. Global South laborers, whose data trains AI models, receive no compensation for their contributions, mirroring colonial extraction patterns. The narrative’s focus on 'productivity' obscures these disparities, framing marginalized workers as passive beneficiaries rather than active agents in shaping economic futures.