The thesis
AI agents compress the cost of collecting information, comparing scenarios, and producing competent analysis. That changes the value of thinking as a standalone advantage: an answer is less scarce when thousands of systems can generate one on demand.
The harder advantage is deciding which question deserves attention, which evidence deserves trust, and which conclusion is important enough to act on.
What remains scarce
Judgment carries responsibility. Trust takes time to earn. Distribution determines whether an idea reaches capital and institutions. Action requires permission, relationships, temperament, and a willingness to live with the result.
Those constraints do not disappear when intelligence becomes abundant. They become more visible.
- Judgment over raw output
- Trust over interface polish
- Distribution over isolated insight
- Disciplined action over endless analysis
Why Buffett still matters
The Buffett lesson was never simply that one person could process more information than the market. It was that temperament, incentives, patience, reputation, and capital allocation compound together.
AI can strengthen the analytical layer. It does not automatically supply the conviction or institutional trust required to deploy capital through a full cycle.