Why AI Is Sucking Up the World's Wealth

The speaker explains that money circulating around the world heads toward the highest expected returns, but is now converging on the AI boom alone, concentrating astronomical amounts of capital in the hands of a few US Big Tech companies, with foreign ownership of US equities reaching 20%. They say that the wealth and savings accumulated by South Korea, Japan, and Taiwan through supplying semiconductors, GPUs, and memory for US AI data centers flow back into US equities and corporate bonds, creating a circular flow that finances the boom and accelerates the AI revolution. They note that in 2025, current-account surpluses and deficits each total 2.5 trillion dollars, marking the largest gap between surplus and deficit countries since 2008, and that, unlike the early 2000s, when funds managed by conservative central banks remained in government bonds, hedge funds, sovereign wealth funds, and pension funds are now flocking to equities. They explain that AI's absorption of capital makes US Treasury securities harder to sell, and that worsening borrowing terms would increase the interest burden. With annual interest payments of 1 trillion dollars on debt of 40 trillion dollars, Trump's goal of reducing the trade deficit clashes with going all in on AI. They say that geopolitical bloc formation, in which countries with similar UN voting patterns invest in one another, has left cross-border investment stagnant since it reached 141% of GDP in 2008, preventing capital from reaching emerging economies that need infrastructure funding. They warn that whether this is the dawn of a new era or a false promise is the trillion-dollar question, and that a sharp fall in valuations would deal Asia a double blow: valuation losses on investments in the US and deterioration in trade balances and economic conditions. They identify the combination of trade imbalances and rising global risk assets as a source of volatility. All directional outlooks are the speaker's conditional judgments, not definitive short-term price predictions.