In a new blog post, the Dutch central bank examines how artificial intelligence (AI) could increase productivity in the Netherlands, drawing on a new study by its economists. The gains depend on businesses adapting their operations, successful applications spreading across the economy and the most productive firms being able to grow. The scale and timing of the economic benefits remain uncertain.
Companies must first invest and reorganize before realizing productivity gains. This can produce a productivity J-curve, in which growth initially weakens before accelerating. Integrating AI into work processes must be accompanied by the movement of labor and capital toward businesses that use the technology most productively.
Europe’s absence from the race to dominate AI development need not prevent it from benefiting, because much of the technology’s economic value comes from its use. However, reliance on U.S. computing power and cloud infrastructure creates risks to access during geopolitical tensions. Unreliable AI systems also pose risks, while competition for market dominance increases the likelihood of harmful incidents in the absence of rules to prevent them.