To effectively incorporate AI into their production processes, companies need to make significant investments in new software, communications, factories, and equipment. However, there is little evidence to suggest a surge in spending in the AI sector.

While some companies have significantly increased capital expenditure to take advantage of AI opportunities, the overall picture remains weak. For instance, forecasters predict that Microsoft’s spending, including research and development, will likely rise by nearly 20%. Nvidia’s capex is also projected to soar by over 30%. However, plans for capital expenditure outside of leading AI companies like Microsoft and Nvidia are more modest.

Investment in ‘information-processing equipment and software’ by American firms fell by 0.4% year on year in the third quarter of 2023, which goes against the expectation of increased spending in this area. Similar trends can be observed globally, with investment spending, including government spending, growing slower than in the pre-pandemic years. This weak capital expenditure growth is leading to limited productivity improvements, as evidenced by a real-time measure derived from surveys of purchasing managers.

These trends suggest two potential interpretations: first, there is speculation that generative AI may struggle to find enough customers for the products and services created through substantial investments by big tech firms. This possibility aligns with previous instances of overestimated demand for new innovations, such as cryptocurrencies and the metaverse. Second, it is plausible that there has been an overestimation of the anticipated demand for generative AI. Regardless of the exact explanation, the evidence indicates limited investment and spending in the AI sector, challenging the widespread excitement surrounding its potential.