Nobel Prize-winning economist Christopher Pissarides has issued a sobering assessment of artificial intelligence's capacity to rejuvenate stagnating Western economies, directly challenging the prevailing narrative promoted by technology leaders and policymakers who view AI as an economic panacea. In interviews and recent public remarks, Pissarides, who specialises in how automation reshapes labour markets, has argued that the era of sustained high productivity growth enjoyed in the 1980s and 1990s may be irretrievable, regardless of AI's technological prowess or adoption rate.
The stakes of this intellectual debate extend well beyond academic circles. For nearly two decades, Western governments and central banks have grown increasingly anxious about sluggish economic performance, particularly across Europe, where productivity growth has remained anaemic compared to historical patterns. This persistent weakness has constrained policymakers' ability to fund public services and manage budgetary pressures while simultaneously dampening real wage growth for ordinary workers. Against this backdrop of economic malaise, both corporate leaders and government officials have invested considerable political capital in the promise that AI will catalyse a return to the robust growth rates that once characterised advanced economies.
Pissarides's scepticism rests on a straightforward observation: concrete evidence of productivity gains from artificial intelligence remains elusive. When asked about claims made by prominent technology figures such as Nvidia Corporation's chief executive Jensen Huang and OpenAI's Sam Altman—both of whom have predicted far-reaching economic and labour market disruption—Pissarides highlighted the absence of measurable proof. This gap between technological capability and actual economic impact deserves particular attention from Malaysian policymakers, many of whom are themselves wrestling with questions about how to position their economies amid rapid AI development globally.
Crucially, Pissarides points to the sectoral composition of modern economies as a fundamental constraint on potential productivity gains. His analysis reveals that approximately 40 percent of jobs in the United Kingdom—and a comparable proportion in the United States—remain largely insulated from AI-driven disruption. Service-sector occupations, particularly nursing and hospitality, depend on human interaction, emotional intelligence, and physical presence in ways that current and foreseeable AI technologies cannot replicate. For such roles, productivity improvements through automation are neither feasible nor practical, meaning large swathes of the labour force will continue operating under existing productivity conditions regardless of advances in machine learning.
During a July 6 lecture delivered at the Royal Economic Society conference in Newcastle, Pissarides elaborated on this structural challenge. Even in sectors most exposed to artificial intelligence—primarily finance and certain professional services—the productivity gains would need to be extraordinary to generate the economic growth rates that technology optimists have forecast. The mathematics are unforgiving. When a sector represents perhaps 10 to 15 percent of total economic output and already operates at relatively high productivity levels, even doubling or tripling efficiency gains would yield modest impacts on aggregate growth. To achieve the transformative expansion some have predicted would require not merely improvement but wholesale economic reorganisation.
The comparison with previous technological waves proves instructive. The computerisation revolution of the late twentieth century fundamentally reshaped how work was organised, how businesses operated, and how information flowed through economies. The subsequent internet revolution enabled new business models and entirely new industries. Pissarides argues that the likelihood of artificial intelligence producing equivalent systemic transformation appears remote, particularly given what is observable about the technology's current capabilities and limitations. His conclusion is intentionally provocative: the days of rapid productivity growth may simply be over, and societies must adjust their expectations accordingly.
This assessment carries significant implications for Southeast Asia and Malaysia specifically. Many regional policymakers have positioned their nations as AI adoption hubs, hoping that early embrace of the technology might leapfrog development stages and accelerate convergence toward advanced economy income levels. Pissarides's analysis suggests such expectations may be misplaced. If even optimally positioned developed economies cannot harness AI for substantial productivity revival, developing nations face even steeper challenges given existing skills gaps, infrastructure constraints, and institutional limitations. The prospect of using AI as a solution to sluggish growth may prove disappointing across both developed and developing contexts.
Not all major voices concur with Pissarides's cautious stance. Andrew Bailey, Governor of the Bank of England, represents the more optimistic camp within policymaking institutions. Bailey has explicitly described artificial intelligence as potentially transformative for economic growth, though he has tempered such remarks by acknowledging that the transition from technological capability to measurable economic impact requires time. Even under his more bullish interpretation, Bailey's position stops short of the triumphalist narratives emanating from Silicon Valley and remains hedged with qualifications about timing and implementation challenges.
The tension between Pissarides's empirically grounded scepticism and official optimism reflects a genuine uncertainty about artificial intelligence's economic future. Pissarides's 2010 Nobel Prize recognised his analytical contributions to understanding labour market dynamics, particularly how friction and information asymmetries affect employment patterns. This established credibility lends weight to his current warnings, though it does not eliminate the possibility that he may underestimate AI's eventual impact—a point he himself acknowledges through careful language about the inherent uncertainties surrounding novel technologies.
For Malaysian policymakers and business leaders monitoring these debates, the practical lesson may be to pursue AI adoption strategies that focus on realistic, sector-specific improvements rather than betting national development strategies on transformative economic acceleration. This more measured approach would involve identifying genuine use cases where artificial intelligence can demonstrably improve productivity or service delivery, while simultaneously preparing workforces and institutions for a future in which AI integration produces incremental rather than revolutionary changes. Such pragmatism might ultimately prove more valuable than either unbridled optimism or blanket scepticism.
