The AI Race: Is the Global Economy at Risk of a $3 Trillion Bubble?

Recent dramatic shifts in the tech world, particularly the massive investments in AI infrastructure, pose both opportunities and serious risks for the global economy. Analysis by David Cahn of Sequoia Capital shows that for the spending by Silicon Valley giants on AI chips and data centers to be justified, the industry must generate $3 trillion in revenue. This is reported by Techcrunch.com reports.

These figures are not mere estimates but are based on NVIDIA's record profits from GPU sales and the rising costs of building data centers. While estimated at around $200 billion in 2023, infrastructure spending is expected to reach $1.5 trillion by 2026. This means that for investors to recoup their capital, revenue must be double that amount.

The gap between revenue and expenditure

Currently, industry leaders like OpenAI and Anthropic are significantly increasing their annual revenue. For instance, while OpenAI reported reaching $20 billion in annual revenue, Anthropic is targeting the $60 billion mark. However, these figures seem small compared to the expected $3 trillion requirement. Tech giants like Google, Meta, Microsoft, and Amazon do not plan to see real returns on these investments until 2028.

Another factor complicating the situation is the rising cost of memory hardware and the increasing demand for complex chips. Construction costs and energy consumption are also becoming more expensive daily, which has sharply increased the revenue required per gigawatt of power.

Risk of economic crisis

Torsten Slok, chief economist at Apollo Global Management, warns that if AI does not deliver the expected financial results, it could deal a blow to the entire global economy. Many organizations are currently shifting from expensive models to cheaper, open-weight models, particularly Chinese developments. This negatively impacts the revenue of major model creators.

Technological efficiency may also yield unexpected results. For example, OpenAI, led by Sam Altman, announced that its new models are 54% more efficient at writing code. While this is good for users, it means financial losses for "token factories" that earn revenue through token sales. If usage volume does not increase sharply, companies may fail to cover their costs.

Experts believe that if these giant companies fail to meet their goals, the S&P 500 index could drop sharply, potentially leading the economy into a recession. The outcome of this tech race is also important for emerging markets, as global economic instability directly affects international investment and technology imports.