Contrary to expectations, global investment in information technologies, a crucial component for AI integration, is not experiencing a surge. In the third quarter of 2023, American firms’ investment in ‘information-processing equipment and software’ actually fell by 0.4% year on year.
Artificial intelligence (AI) has long been hailed as the transformative force poised to revolutionise the global economy. Economists and industry experts have anticipated an era of explosive growth, with AI potentially substituting for human labour and driving significant increases in GDP. However, as we examine the current state of affairs, it becomes apparent that the much-anticipated AI investment boom has not materialised as expected.
Capex Trends and Slowdowns
Traditionally, major technological breakthroughs, such as the tractor or personal computer, necessitated significant investments in new software, communication infrastructure, factories, and equipment. These investments facilitated the integration of ground-breaking technologies into production processes, resulting in widespread economic impact. In the case of AI, a similar investment boom was expected to pave the way for its transformative potential. Yet, the reality remains stark – there is little evidence of a substantial AI splurge in capital expenditure (capex) by businesses globally.
After a period of sluggish growth preceding the COVID-19 pandemic, global capex increased with the easing of lockdowns. By early 2022, it was rising at an annualised rate of approximately 8%. However, this momentum was short-lived, slowing down later that year due to geopolitical uncertainty and higher interest rates. By March 2023, on the eve of the release of OpenAI’s GPT-4, global capex spending was growing at an annualised rate of about 3%. Today, while some companies, particularly tech giants like Microsoft and Nvidia, are increasing capex to seize AI opportunities, the broader business landscape appears less enthusiastic.
Microsoft and Nvidia, at the forefront of the AI revolution, are forecasted to experience significant capex increases, reflecting their commitment to AI development. However, when excluding these industry leaders, S&P 500 companies plan only a modest capex lift of around 2.5% in 2024—barely surpassing the rate of inflation. The overall economic scenario is even more disheartening, with an American capex “tracker” from Goldman Sachs revealing a 4% year-on-year decline. These figures paint a less optimistic picture of the broader adoption of AI across industries.
Global Trends in AI Investment
Contrary to expectations, global investment in information technologies, a crucial component for AI integration, is not experiencing a surge. In the third quarter of 2023, American firms’ investment in “information-processing equipment and software” actually fell by 0.4% year on year. This trend is not confined to the United States, as national-accounts data for OECD countries indicate slower investment spending growth compared to pre-pandemic years. A high-frequency measure of global capex from JPMorgan Chase points to minimal growth, contributing to a lack of productivity improvements.
Regional variations further complicate the AI investment landscape. Japan, for instance, shows a promising increase in capex growth, potentially fuelled by specific factors like corporate governance reforms. However, outside America, the situation is less encouraging. A challenging economic outlook in Europe has dampened investment intentions, with services companies in the European Union planning less ambitious capex increases than in early 2022. British businesses, too, exhibit a more conservative approach, planning a mere 3% increase in capex over the next year, significantly down from the 10% projected in early 2022.
These trends prompt two plausible interpretations of the current state of AI investment. The first suggests that generative AI may be a disappointment, with big tech firms struggling to find customers for products and services despite substantial investments. This scenario draws parallels with recent history where technologists overestimated demand for innovations like cryptocurrencies and the metaverse.
The second, more optimistic interpretation aligns with historical patterns in the adoption of general-purpose technologies. The analogy with the personal computer’s trajectory is apt, where despite a ground-breaking operating system release by Microsoft in 1995, substantial spending on software only occurred in the late 1990s. Analysis by Goldman Sachs indicates that while only 5% of chief executives expect AI to have a “significant impact” on their business within one to two years, a substantial 65% foresee an impact in the next three to five years. This interpretation suggests that AI is likely to transform the economy, but the process will unfold gradually, more with a whimper than a bang.
The anticipated AI investment boom appears elusive at present, raising questions about the transformative potential of generative AI in the immediate future. While some industry leaders are actively increasing capex to capitalise on AI opportunities, the broader economic landscape remains restrained. Historical patterns in the adoption of transformative technologies and the cautious optimism expressed by business leaders suggest that AI’s impact may indeed be profound but will require patience as businesses gradually integrate this revolutionary technology into their operations.