Economic Resiliency Amid Global Challenges: The Role of AI

The world economy has been more resilient than many expected in 2024. That is encouraging, but growth figures alone do not settle the questions raised by geopolitical tensions, debt and uneven productivity. I am particularly interested in how AI could affect that last part.

The IMF’s April 2024 outlook projected global growth of 3.2% in 2024 and 2025. The forecast combined continued expansion with substantial differences between countries. Inflation, public debt and geopolitical tensions remained risks.

Trade restrictions and competition over advanced technology also complicate the outlook. Those tensions matter for AI because access to computing infrastructure and data affects who can develop and use it.

Who captures the productivity gains?

AI can reduce the time required for some tasks while changing the skills needed for others. A firm may produce more with the same team, expand its activity, or reduce hiring. Those choices have different consequences for workers, even when each improves the firm's productivity.

I would look at the transition as closely as the eventual gain. People need time and support to learn new work. A training programme is useful when it leads to tasks employers actually need, and when workers can afford the time it takes to learn them.

Financial markets introduce another uncertainty. Expectations of future AI profits can move faster than evidence from customers. I would distinguish a company showing a measurable improvement in its operations from one whose valuation mainly reflects the promise of that improvement.

Access to chips, computing infrastructure and skills also affects who can participate. Concentrating those resources in a few firms or countries could concentrate the gains. Smaller businesses need affordable ways to use the technology, while retaining enough control over their data and workflows to change providers.

Cross-border cooperation could spread capabilities more widely. Trade restrictions and competing national interests could instead make access more fragmented. I see both possibilities, and neither follows automatically from better models.

The aggregate growth figure is only the beginning of the story. I want to know who becomes more productive, who receives the benefit, and what happens to the people whose work changes before a new opportunity appears.