Christopher Pissarides, the Nobel Prize-winning economist specialising in labour market dynamics and workplace automation, has delivered a sobering assessment of artificial intelligence's economic potential. Speaking to Bloomberg News, Pissarides argued that AI will not resurrect the era of brisk productivity expansion that characterised Western economies decades ago—and that such rapid growth may have become permanently unattainable regardless of technological breakthroughs.
The London School of Economics professor's scepticism stands in sharp contrast to the euphoria surrounding AI in corporate and policy circles. Technology leaders like Nvidia Corporation's Jensen Huang and OpenAI's Sam Altman have promoted the narrative that artificial intelligence represents a transformative force capable of fundamentally restructuring economic output and employment. Yet Pissarides, who won the Nobel Prize in Economics in 2010 for his groundbreaking research on labour market frictions, contends that this optimism rests on flimsy empirical foundations.
A crucial constraint identified by Pissarides is the limited exposure of significant portions of the workforce to AI-driven productivity improvements. He estimates that approximately 40 percent of jobs in the United Kingdom—and similarly substantial proportions in the United States—will remain largely insulated from artificial intelligence's direct effects. Sectors such as nursing, hospitality, and other service industries dependent on human interaction and physical presence cannot easily be automated or augmented by current AI capabilities, meaning these segments will not experience the productivity windfall that technology enthusiasts anticipate.
The underlying concern here reflects broader economic realities facing Western nations. Over recent decades, productivity growth has decelerated markedly across developed economies, particularly in Europe. This sluggish expansion has constrained governments' fiscal options and complicated policymaking, as resources for addressing social needs have become scarcer. Simultaneously, wage gains for ordinary workers have stagnated in real terms, contributing to economic anxiety and political turbulence across multiple democracies. The hope that AI might revitalise growth and ease these policy tensions has therefore become an article of faith for many political and corporate leaders.
Pissarides' analysis demands a fundamental reckoning with these hopes. Despite the technological breakthroughs of the past eighteen months, he observes little concrete evidence that AI has begun delivering meaningful productivity improvements across the broader economy. Current metrics show no measurable acceleration in output per worker or other standard productivity measures. This absence of evidence is particularly telling given the enormous investment in AI infrastructure and adoption that has already occurred.
During a lecture delivered on July 6 at the Royal Economic Society conference held in Newcastle, Pissarides elaborated on his thesis. Even if artificial intelligence were to generate significant productivity gains in the most exposed sectors—particularly finance and certain professional services—the overall boost to the economy would remain underwhelming. The mathematics simply do not support the bullish scenarios: because so many workers operate in AI-resistant sectors, even large productivity gains in finance or technology cannot translate into the sweeping economic acceleration that optimists have projected.
The economist explicitly rejected comparisons to previous technological revolutions. The computer revolution of the 1980s and 1990s, which fundamentally reshaped business operations and created entirely new industries, remains the standard against which AI proponents benchmark their expectations. Yet Pissarides expressed profound doubt that artificial intelligence will prove similarly transformative. "Given what we know now and what we see happening, I don't see the productivity growth matching those levels," he stated, while acknowledging genuine uncertainty about technology's trajectory.
His conclusion borders on resigned acceptance. The era of rapid productivity growth that powered Western prosperity from the post-war period through the early 2000s may simply have ended. This represents not a temporary cyclical downturn but a structural shift in the economy's growth potential. AI may provide modest improvements at the margins, yet these incremental gains cannot compensate for the underlying slowdown. "It's just not practical to talk about high productivity growth," Pissarides argued. "I think we should be resigned to the fact that the days of fast productivity growth are over, whatever we do."
For policymakers in Southeast Asia and Malaysia specifically, this assessment carries important implications. Many regional governments have embraced AI as a development accelerator, hoping the technology will enable catch-up growth and technological leapfrogging. If Pissarides is correct that AI's genuine economic payoff remains modest even in advanced economies with greater AI adoption, then developing nations cannot reasonably expect artificial intelligence alone to overcome the structural constraints limiting their growth. Instead, sustained advancement will require continued investment in human capital, institutional quality, and the unglamorous fundamentals of economic development.
Yet not all policymakers have surrendered to pessimism. Bank of England Governor Andrew Bailey has positioned AI as potentially transformative for growth, though he cautioned that meaningful effects will take time to materialise in official statistics. Bailey suggested the technology "may well ride to the rescue" of sluggish Western economies. This divergence between Pissarides' scepticism and Bailey's cautious optimism reflects genuine uncertainty about artificial intelligence's economic impact, a question that will ultimately be settled by empirical observation rather than rhetorical assertion.
The broader lesson from Pissarides' intervention concerns the limits of technological determinism in economic policymaking. While new technologies can reshape production and employment, they operate within structural constraints that technology alone cannot overcome. The challenge for policymakers across the developed and developing world is to build resilient, adaptive economies capable of delivering broad-based prosperity even when headline productivity growth remains modest—a considerably more demanding task than simply betting on artificial intelligence to revive the rapid expansion of earlier decades.
