I was there for the dot-com burst. Here's how the AI bubble will pop.

In the early stages of new technologies, intense debates often arise regarding whether the spending surge is a rational investment or a speculative bubble. However, AI is poised to significantly alter the economy and society, just like the internet, computers, cars, trains, canals, electricity, and other innovations have. The money being poured into AI right now is likely to be lost, just as much of the funds invested during previous booms were.

The early AI boom may conclude with a massive bust that wipes out the initial wave of companies and investors, followed by a prolonged period of growth that creates substantial new firms and transforms daily life. I sincerely hope an AI bust never occurs, as it would devastate millions, including myself. While I do not operate an AI firm or trade frequently, I am still invested in stocks and depend on a thriving economy.

Therefore, if an AI bust were to take place, I would be adversely affected, along with everyone else. However, having witnessed two major booms and busts - the Internet (1995-2002) and the Great Financial Crisis (2002-2009) - I find the parallels and differences intriguing.

Much like their predecessors, the AI boom is characterized by growth and leverage. Growth stems from an innovative product or service that generates immense demand, while leverage amplifies this demand through debt, credit, circular financing, or other means. In the internet boom, real demand rose as millions of new users connected monthly, leading to astonishing growth for numerous companies.

The leverage was evident, particularly when companies borrowed billions to develop telecommunications networks and acquire new equipment. Similarly, AI companies are experiencing unparalleled growth, with revenue exploding for some leading firms like Anthropic. In the second quarter, Anthropic's revenue doubled to $11.6 billion, and it now generates an astonishing $65 billion in annualized revenue.

This growth rate is unprecedented, and some analysts predict Anthropic will generate $100 billion in revenue by early next year.

However, just as with the internet and the Great Financial Crisis, a significant challenge lies ahead. If the supply of AI products and services catches up with demand, and financing leverage peaks, the AI boom may face a bust. The global market for enterprise software spending, the primary source of revenue for AI companies, is estimated at $1.4 trillion, providing ample room for expansion. Still, the extent to which AI can continue its current trajectory remains uncertain.