AI Infrastructure Investment Heats Up

Investors are increasingly focused on the infrastructure underpinning the artificial intelligence boom, but should look beyond headline figures like processing power. The real value lies in controlling essential inputs that are difficult to commoditise. Demand for data storage and computing power is surging, requiring massive investment in data centres, electricity, and fibre networks. AI implementation is also improving efficiency and optimising operations across various infrastructure assets, which could translate to enhanced returns for investors.

AI infrastructure spending is projected to reach $758 billion by 2029, with cloud and shared environments currently dominating deployments. Hyperscalers and cloud service providers are the biggest spenders, particularly on accelerated servers. However, concerns exist regarding overinvestment and shrinking returns, as historical infrastructure booms have often led to excess competition and poor stock performance.

To justify the current levels of investment, AI revenues need to increase dramatically. Investors should, therefore, maintain a degree of scepticism about infrastructure builders' ability to effectively monetise their investments at current valuations.