AI race to transform the world before the money runs out

The AI industry is experiencing an unprecedented financial surge, dwarfing previous technological revolutions like railways and the internet. According to projections by PwC, global spending on data centers could reach $30 trillion by 2050, nearly matching the value of US Treasuries. Major companies, including Anthropic, are investing staggering amounts in the AI race.

For instance, Anthropic plans to spend $518 billion in the coming years, more than 100 times its 2025 revenue. However, the projections and investments are based on assumptions about significant productivity gains and future profits, which lack historical evidence or precedent. Economists question whether broad-based productivity gains in the US, the leading AI nation, are truly achievable, given the elusive nature of such gains so far.

While AI is expected to transform various sectors, the question remains whether these advancements will be sufficient to fund the massive investments. US hyperscalers like Google, Amazon, and Microsoft, along with other players in the AI race, will need to generate over $4.2 trillion in new revenue within five years to finance their expansions.

JP Morgan estimates that US productivity gains would need to be 3% to 5% annually over the next decade to justify the valuations of companies like Nvidia, whose chips are crucial to the AI revolution. Analysts warn that a downturn in demand, delays, or asset devaluation could result in significant losses due to the leveraged structure of AI infrastructure debt.

Despite the enormous sums at stake, AI leaders remain optimistic, with some envisioning a future of "transcendent beauty" and unprecedented productivity gains through self-teaching AI models.