AI Investment Boom: History's Warning on Tech Bubbles

Summary

History's capital cycles, from railroads to the dot-com era, offer stark lessons for today's AI investment frenzy, according to financial historian Edward Chancellor. He explains that while new technologies attract immense capital, investors often misjudge the speed and scale of profits, leading to overinvestment and eventual shakeouts. Examples include the railway mania, the proliferation of car and aircraft companies, and the dot-com bust, where despite eventual winners like Amazon, massive losses occurred. Chancellor highlights that markets struggle to spot winners, and the AI boom might see similar pitfalls, especially if demand is overestimated or if inherent flaws in technologies like large language models limit their applicability. He points to historical instances where excessive investment in physical capital like fiber optic cables, while eventually enabling new services, led to significant economic disruption. Furthermore, Chancellor notes that intangible capital, such as R&D and brand value, is also susceptible to boom-and-bust cycles, citing pharmaceutical R&D and the SaaS bubble. He cautions against the idea that bubbles are inherently productive for investors, emphasizing the risk of capital misallocation and the potential for severe long-term consequences, like the global financial crisis stemming from the dot-com bust. Looking at current markets, Chancellor suggests seeking 'anti-bubbles' – neglected sectors that may be unfairly discounted due to AI disruption fears, while cautioning against companies that might be genuinely outmoded. He also maintains a long-term positive outlook on gold as a portfolio hedge.