US Senator Elizabeth Warren has issued a warning about the financial risks of the current artificial intelligence boom. At an event in Washington, she compared the situation to the precedents of the 2008 crisis, pointing to practices of uncontrolled spending and massive borrowing. AI companies are being funded with opaque funds and little oversight. If they don't generate real revenue soon, they could collapse, dragging down the system.
The Hidden Cost of Development: Debt, Chips, and Promises 📈
The model is simple and worrying. Companies are consuming capital at a frenetic pace to pay for computing infrastructure, like GPU clusters, and specialized talent. This funding largely comes from private credit funds, a market with less transparency than traditional banking. The pressure to scale larger models competes with the need to find sustainable business models. Debt accumulates before products demonstrate profitability at scale.
2008: Mortgage Subprime. 2024: Parameter Subprime ⚠️
History repeats itself, but with a science fiction script. Before it was packages of junk mortgages, now they are packages of debt backed by promises of AGI and tokens. Private credit funds are the new big players, making colossal bets off the radar. All it takes is for one flagship company to stumble and panic to spread faster than a language model generating excuses. At least this time the bubble has a more modern name.