Everyone asks whether the AI bubble will burst. Does your AI strategy survive both scenarios?
Everyone asks whether the AI bubble will burst.I ask something else: does your AI strategy survive both scenarios?
If the bubble bursts, the market splits in two.
Compute gets cheap: written-off hardware, bankrupt operators, capacity sold at marginal cost. But the frontier labs lose their subsidy. They sit on compute commitments in the hundreds of billions and need margins for the first time. The survivors consolidate and their token prices rise. Cheap raw material, expensive finished product.
If it holds, the market consolidates anyway. The winners raise prices, rewrite terms, deepen the lock-in.
Either way, you end up in front of the same question: how much of your AI do you actually control?
These three questions help. Could you switch the model? Could you run one yourself on that cheap compute if you had to? Do you know where your data flows, and under whose law?
Most companies answer none of them. Because at today’s prices, the questions feel unnecessary.That feeling is exactly what the price is buying.
A strategy that needs the bubble question answered first was never a strategy.
Related posts
- Klarna and Duolingo reversed AI-first before the EU adopted itKlarna rehired people 18 months after cutting 700 support roles. Eight accountability questions to settle before an AI procurement decision.
- 41% of Y Combinator's AI bets sit where nobody wants themA Stanford matrix of human desirability against technical capability. In 45.2% of occupations workers want equal partnership, not a replacement.
- If the AI bubble bursts, Europe falls softer. That is a warning, not an advantageEurope has invested less in AI and would lose less in a crash. Many call that an advantage. Marc reads it as a reminder of the year 2000.