AI Governance and Global Inequality Explained

AI Governance and Inequality

Syllabus

GS 3: Artificial Intelligence

Why in the News?

Recently, AI pioneer Geoffrey Hinton warned that Artificial Intelligence could enrich a few while impoverishing many, reviving debates on a modern “Engels’ pause” marked by rising productivity but stagnant wages.

Introduction

• Artificial Intelligence (AI) has emerged as a transformative force in the global economy.
• Yet, Nobel Laureate Geoffrey Hinton recently warned that AI may make a few people rich while leaving most others poorer.
• His statement echoes a historical paradox called “Engels’ pause”, raising concerns about unequal gains from technological progress.

Understanding Engels’ Pause

• The term “Engels’ pause” refers to an economic puzzle in 19th century Britain.
• Industrial output grew rapidly, but ordinary living standards did not improve for decades.
• Wages stagnated, food absorbed most household budgets, and inequality widened.
• Oxford economist Robert Allen coined the term in a paper, naming it after German philosopher Friedrich Engels, who had observed these inequalities.
• Despite Britain becoming the “workshop of the world,” only much later did ordinary citizens see welfare improvements.

Why This Idea Matters Today

• As AI reshapes industries, a similar paradox may emerge: productivity gains without broad-based prosperity.
• The question today is whether we are entering a modern Engels’ pause.
• The warning becomes sharper with evidence such as:
• A Stanford study showing younger workers are more vulnerable to AI-related disruptions.
• An Indian software giant cutting 12,000 jobs while pivoting to AI.
• A MIT study revealing that 95% of AI pilots fail to produce visible organisational benefits due to missing complementary skills and systems.
• These trends show that while AI may increase efficiency, workers and households may not share in the benefits immediately.

AI as a General-Purpose Technology (GPT)

• Economists describe AI as a general-purpose technology, similar to steam power, electricity, or the Internet.
• GPTs are unique because they:
  • Transform multiple industries.
  • Lower the cost of core functions (AI lowers the cost of prediction, as noted by Agrawal, Gans, and Goldfarb in 2018).
  • Trigger new opportunities but also cause dislocation.
• Historical studies by Nicholas Crafts (2021) and Bojan Jovanović with coauthors (2005) show that such technologies initially benefit capital owners and entrepreneurs more than ordinary workers.
• AI may repeat this pattern, enriching a few while leaving many behind, exactly as Geoffrey Hinton cautioned.

Markers of a Modern Engels’ Pause

Productivity Gains but Stagnant Wages

• In Philippine call centres, generative AI copilots have raised productivity by 30–50%, reducing costs for companies and speeding up service.
• However, worker wages have not increased proportionally, and workloads have sometimes intensified.
• This mirrors the 19th-century paradox where industrial growth coexisted with wage stagnation.
• At the same time, high inflation and rising living costs make workers feel poorer despite productivity improvements.
• A recent New Yorker cartoon humorously highlighted this paradox, showing a person asking ChatGPT why her electricity bills are rising — a reflection of AI’s darker side for households.

Rising Costs of Complements

• AI productivity depends on expensive complements such as:
  • Cloud computing
  • Cybersecurity
  • Data access
  • Worker retraining and certifications
• For employees, the “price of staying relevant” has increased. Coding bootcamps, new certifications, and continuous learning now demand significant investment.
• This is similar to how 19th-century wage gains were offset by rising food prices.
• Today, digital survival costs erode modest income improvements.

Unequal Distribution of Gains

• PwC estimates AI could add $15.7 trillion to global GDP by 2030.
• However, gains are concentrated in:
  • The S.
  • China
• A handful of firms controlling foundational models.
• The IMF (2024) estimates 40% of jobs worldwide are exposed to AI.
• Half of these are in advanced economies, where high-skilled substitution is more likely.
• This bifurcation delays or denies welfare improvements for many workers in developing economies.
• Evidence from this writer’s study in the Journal of Development Economics showed that in India, stronger intellectual property laws created deep wage inequality during a technology race a possible preview of the global future.

Job Displacement and Task Transformation

• AI is not just boosting productivity but also reshaping jobs.
• Examples include:
  • Doctors increasingly using ChatGPT as a complement.
  • Researchers at Tsinghua University in China launching the world’s first AI-powered hospital.
  • AI adoption in education, finance, public management, and infrastructure steadily transforming roles.
  • The government of Albania appointing the world’s first AI Minister, Diella.
  • Consulting studies, including recent work with GMR Airports, show AI displacing traditional roles.
• These examples highlight how AI-driven changes may displace workers before welfare gains spread widely.

Historical Lessons

• During the Gilded Age in the U.S., productivity surged but inequality rose sharply.
• Labour unrest and political upheaval followed.
• Only with reforms such as:
  • Trade unions
  • Public schooling
  • Welfare systems
    did living standards rise broadly.
• The lesson: without governance, technological gains may remain concentrated.

Policy Directions to Avoid an AI Engels’ Pause

Skills Transition and Human Capital

• Governments must prioritise reskilling to ensure workers benefit from AI.
• Models include:
  • Singapore’s SkillsFuture programme, which offers continuous education credits for lifelong learning.
  • Mohamed bin Zayed University of Artificial Intelligence (MBZUAI) in Abu Dhabi, the world’s first AI university.
• These initiatives show how systematic investment in skills can support inclusive AI transitions.

Redistribution of AI Gains

• Policies must ensure AI’s economic rents are shared fairly. Options include:
• Robot taxes to redistribute automation profits.
• Universal Basic Income (UBI) schemes, experimented with in the K. and the European Union.
• Philanthropic commitments, such as the Chan-Zuckerberg Initiative, that channel AI wealth toward social good.

AI Infrastructure as a Public Good

• AI productivity depends on compute power and data — the “food” of AI.
• If these remain scarce and expensive, welfare gains will lag.
• Positive steps include:
  • ai (UAE) and Apertus (Switzerland), both launched as public open AI reasoning models in September 2025.
• Treating AI infrastructure as public rather than private ensures broader access to benefits.

Challenge Ahead

• Critics argue the Engels’ pause analogy may be exaggerated.
• Unlike the 19th century, today’s societies have:
  • Stronger welfare systems
  • Democratic institutions (though many face democratic backsliding)
  • Faster technology diffusion smartphones reached billions within a decade.
• AI has potential to lower costs in healthcare, education, and clean energy, offering immediate welfare improvements.
• This suggests today’s pause could be shorter if policy and governance keep pace.

Balancing Optimism with Caution

• We must guard against a situation where macro-level gains coexist with micro-level stagnation.
• Political economy shows that inequality is not inevitable; it is shaped by political will and governance choices.
• The challenge for policymakers and AI governance experts is to ensure that AI becomes not only a productivity revolution but also a human welfare revolution.
• History reminds us: progress delayed is progress denied.
• Whether this pause lingers or passes quickly depends on policy action and social choices.

Conclusion

AI has the potential to be as transformative as steam power or electricity, but without governance, it risks triggering a modern Engels’ pause. Productivity may rise, but wages and welfare gains may stall. Whether AI uplifts all or only a few will depend on choices made today.

Source: The Hindu

Mains Practice Question

Discuss how Artificial Intelligence, as a General-Purpose Technology, can create both productivity gains and economic dislocations. What policy interventions are necessary to prevent an “AI Engels’ pause”?