Power in the Age of Intelligence

Author: Packy McCormick URL: https://www.notboring.co/p/power-in-the-age-of-intelligence One-line: When technology gets abundant, the returns stop going to the people who own clever software and start going to the people who identify an industry’s binding constraint, break it, and integrate outward from the position that breaking it creates.

The argument

McCormick’s claim is that the standard moat vocabulary — the checklist an analyst runs on a company — was built for a period of relative technological stability and is actively misleading in a period of rapid change. His borrowed frame is military: a Schwerpunkt, the point of main effort where an attack is concentrated, and the High Ground you occupy once you have broken through it. The investing translation: find the constraint that governs an entire industry, break it, take the scarce position that breaking it creates, then expand outward in every direction the position allows.

His evidence is the last industrial revolution rather than this one, which is the essay’s best structural choice — it means the pattern is not being read off the very companies it is meant to predict.

• Rockefeller won by breaking the refining constraint and then integrating outward, horizontally through acquisition of struggling refiners and vertically into distribution, with scale and efficiency mattering more to the outcome than the famous railroad rebates.
• Gustavus Swift could not ship dressed beef because no cold chain existed, so he built the whole thing himself — refrigerated cars the railroads refused to pull, so he leased his own; icing stations along the route, supplied by contracts with Wisconsin ice harvesters. Vertical integration by necessity, and then a durable position once assembled.
• Ford went from 12,000 cars in 1909 to half a million in 1916 to over two million in 1923, with more than half the cars in the world at peak.

The point of the sequence is a challenge to the reader: would any moat checklist of the era have identified these people in advance, let alone told you how to become one? McCormick’s answer is no, and his prescription is that this kind of underwriting cannot be done in a spreadsheet — it requires reasoning about constraints rather than scoring attributes. He leans on the Teece argument that returns often accrue to the owners of complementary assets rather than to the developer of the intellectual property, which is why the innovating firm needs a prior position in those assets.

The contemporary application is where the essay earns its keep. He reports founders being pushed by investors toward selling software, because software revenue is high-margin, fast and legible to downstream capital — and argues that in an era of software abundance, a point-solution software company faces existential risk from the very progress it is riding. Hardware is not automatically the answer either. The question is always which position an industry’s constraint makes scarce.

His worked example is Base Power Company, which he treats as the template: buy commodity cells, manufacture packs, install them on homes starting in Texas, write the coordination software, trade in the power market, and partner with utilities to balance the grid. The logic chain is explicit — we want to fix power, the bottleneck is the grid, so own the position that relieves the bottleneck and expand from there. He adds Somos (starting in a high-need, regulation-friendly market and building cash before bigger fights), Astro, and Earth AI, whose story is the cleanest illustration: it could not learn from mineral-exploration customers because they were too slow to adopt, so it bought its own rig, discovered it could build better rigs, and simply became the competitor. The same slowness that made them bad customers made them attractive targets.

The essay’s asymmetry argument is the part worth carrying around: improvement compounds unevenly. Better software is worth more to Base than to a smaller competitor or a pure generation company, better robots make it faster and more profitable, and making the game more capital-intensive actively disadvantages would-be entrants. He closes on valuation as evidence that the market already prices this — Ramp against Brex, Stripe against Adyen, Base at $4bn two years in — because owning the scarce position in an industry is itself the scarce asset, and from it you can eat an industry.

Notable quotes

Innovation alone, software or hardware, isn’t enough.

The ownership of the scarce position in an industry is itself a scarce asset.

If you’re just selling point solution software, then software abundance is a threat.

Archived text

Source page saved locally against link rot, with a section-by-section record and the charts: ../attachments/power-in-the-age-of-intelligence/power-in-the-age-of-intelligence.md

Connections

• Packy McCormick — the author, and Base Power is a Not Boring Capital position, which he discloses; the essay is partly a public statement of his underwriting method.
• Venture Capital Models — the essay is an argument that the prevailing model underwrites legibility (fast high-margin software revenue) rather than position, and that this is now a systematic error.
• Vertical Integration — Swift, Ford and Base are all cases where integration was forced by the absence of a working supply chain and then became the durable advantage.
• Moats — McCormick’s direct target. His claim is not that moats do not exist but that the checklist form of the concept is backward-looking and useless during rapid change.
• John D. Rockefeller — the founding case: break the constraint, then integrate outward horizontally and vertically.
• Henry Ford — the scale case, with the 1909–1923 production figures.
• Andrew Carnegie — named in the same sequence of industrial-revolution operators who found and broke a constraint.
• Ramp — his valuation evidence that the market prices the ability to eat adjacent categories, not the current product.
• Aggregation Theory — the useful contrast. Aggregation explains power accruing to whoever owns demand; McCormick is arguing that in capital-intensive physical industries power accrues to whoever owns the relieved constraint.
• Elon Musk and SpaceX — his most extreme illustration of integrating outward from a High Ground.