Nobel Economist: The Era of Fast AI-Driven Productivity Growth Won't Happen

Christopher Pissarides, the 2010 Nobel laureate in economics, said plainly that companies and investors should stop counting on a return to rapid economic growth driven by artificial intelligence. Speaking at a lecture at the Royal Economic Society conference in Newcastle, he said the data economists currently have does not support such expectations.

Pissarides specializes in studying the impact of automation on the labor market, and it is from that vantage point that he assessed the current wave of enthusiasm around language models. According to him, matching the scale of productivity growth seen during the computer revolution of the 1980s and 1990s would require sectors most exposed to AI, such as finance, to post enormous efficiency gains. So far, nothing of the sort is happening.

What the economist actually said

Given what we know now and what we see happening, I don't see the productivity growth matching those levels - Christopher Pissarides, professor at the London School of Economics

I think we should be resigned to the fact that the days of fast productivity growth are over, whatever we do - Christopher Pissarides, professor at the London School of Economics

The economist stressed that this disappointment doesn't mean AI is useless. In his view, large parts of the labor market simply won't feel its impact the way the tech industry's most enthusiastic forecasts suggest. He pointed specifically to personal services and healthcare as sectors where automation runs into natural barriers tied to physical human presence and interpersonal relationships.

Pissarides isn't alone in tempering expectations

Pissarides's view fits into a broader current of skepticism among labor economists. Daron Acemoglu, the 2024 Nobel laureate and MIT professor, previously estimated that AI would automate around 5 percent of all tasks performed in the economy over the coming decade, translating into global GDP growth of only about 1 percent. That is a fraction of what the most optimistic Wall Street forecasts assume.

Acemoglu has also argued that AI's real strength lies not in replacing people but in working alongside them, and that the barrier to breakthrough applications remains so-called tacit knowledge, the informal expertise that is hard to learn from data. Occupations requiring social interaction, complex judgment, or operating in unpredictable environments still lie beyond the reach of current models.

A gap between hype and data

The claims made by both economists clash with the narrative pushed by industry leaders such as Nvidia chief Jensen Huang and OpenAI's Sam Altman, who regularly describe AI as a force capable of rebuilding entire sectors of the economy within a few years. Meanwhile, research from labor market organizations points to a growing gap between companies' declared enthusiasm and the actual results of their AI rollouts.

Data from the National Bureau of Economic Research shows that while one in three surveyed executives say they use AI at work, the average actual usage time is just an hour and a half per week. Analysis from MIT Sloan Management Review points to a classic productivity paradox familiar from earlier waves of digitization, where gains felt by individual workers are partly offset by implementation costs, process reorganization, and the need to adapt the entire organization to the new technology.

What this means for Polish companies

For companies in Poland that have been investing heavily in AI tools in recent months, the remarks from Pissarides and Acemoglu are a signal for caution. Rather than treating the rollout of chatbots or coding assistants as a guarantee of a leap in productivity, it's worth expecting that the real effects will unfold over time and vary sharply by industry.

Service sectors built on direct customer contact, such as healthcare, food service, and hospitality, may see changes far later and to a lesser degree than analytical or IT industries. That means companies' investment strategies should be tailored individually rather than copied from broad declarations about an AI revolution.

Pissarides made clear that his assessment is not a call to abandon investment in technology, but an appeal for realistic expectations. In his view, the labor market will change gradually rather than abruptly, and economic policymakers should plan employment policy around this drawn-out process rather than a vision of sudden transformation.

Sources: The Star (thestar.com.my), MIT Sloan Management Review Poland (mitsmr.pl)