Preface: The Return
This series begins with a claim that sounds contradictory until you sit with it for a while:
The AI bubble will not burst because AI is fake. It will burst because AI is real.
I do not mean the technology is weak. I mean it may be strong enough to redraw the map faster than the market can price it correctly.
That is what real paradigm shifts do. They make the future feel obvious before the ownership layer is settled. Capital rushes toward the most visible winners, but the mature economy often forms somewhere else.
That is why this series starts with dotcom memory rather than a valuation chart. I care less about whether AI is “overhyped” in the abstract than about whether the market is confusing the gateway with the destination, the early winners with the durable winners, and the first story with the final structure.
This is the argument running through every chapter:
- AI utility is already real.
- The first ownership story is probably not the final one.
- A crash, if it comes, may be a repricing, not a negation.
- The real AI economy may emerge after the obvious trade breaks.
Stanford’s AI Index captures how economically and socially present AI already is, even while the long-term power structure remains unsettled.
Series journey
AI and the Programmable Firm
Read this sequence in order. Each chapter advances the same argument from a different angle, building from dotcom memory toward the mature AI power map.
The thesis, the reading contract, and the journey through the full eight-part argument.
Dotcom memory, paradigm shifts, and why real technologies produce the most dangerous bubbles.
Early companies and interfaces are prototypes of futures they may not own. A prototype can be directionally right and economically doomed.
Intelligence alone is not the moat. Durable moats may be context, distribution, workflow, identity, permissions, governance, execution rights, data and systems of record.
SaaS won because custom software was too expensive. AI changes that cost curve, and the compromise gets repriced.
Firm-ware is the operating logic embedded into the company itself. AI makes that logic programmable.
The market does not stop believing in AI. It starts doubting who owns it. The pressure map showing how multiple assumptions weaken at the same time.
The crash is a repricing event. The mature AI economy begins after the first ownership story fails.
A layered view of the durable winners, pressured groups, and control points in the mature AI economy.
How to read the series
Each chapter works on its own, but this is designed as a sequential essay.
Read straight through and the structure should feel intentional:
- Part 1 returns to dotcom memory and the emotional pattern of paradigm-shift bubbles.
- Part 2 states the core claim directly, AI is useful, and that is exactly why the bubble is dangerous.
- Part 3 asks whether intelligence delivered as an API is a durable moat or only an early distribution phase.
- Part 4 argues that SaaS benefited from the old cost of building software, and AI changes that cost curve.
- Part 5 explores the AI-operated company as a possible new organizational form.
- Part 6 maps the pressures that could cause belief to break.
- Part 7 asks what becomes more real after the crash, not what disappears.
- Part 8 ends with the likely power map, who is strengthened, who is pressured, and where durable control may live.
Start with Part 1
If you want the series in the intended order, begin with Part 1: The Business Cards in the Drawer.
That opening chapter is where the memory, mood, and pattern first lock together. Everything after that builds from the same intuition:
the technology can be real, the excitement can be justified, and the market can still be pricing the wrong future.
If that premise interests you, the journey map above is your way in.
If you prefer the ebook option (lacking the interactive sections), then you have the following available: