Notes
The New Scarcities
On this page
When a thing becomes abundant, the value it used to carry does not evaporate — it moves. The phrase "too cheap to meter," borrowed from a 1950s promise about nuclear electricity, is now aimed at intelligence: if cognitive work can be summoned on demand at near-zero marginal cost, the bottleneck simply shifts to whatever intelligence cannot manufacture. Abundance does not abolish scarcity. It relocates it.
Abundance relocates, it does not end
Economics has only ever been the study of allocation under scarcity, and nothing about cheap intelligence repeals that. When one input collapses in price, demand and attention pile onto the inputs that remain stubborn. Cheap thinking makes the non-thinkable parts of the world — the physical, the social, the embodied — the new chokepoints. The useful question stops being "what will AI make plentiful?" and becomes "what stays hard to copy when copying minds is easy?"
Energy and matter stay stubborn
The most concrete floor is physical. An idea can be duplicated for free; a megawatt cannot, and neither can a hectare of land, a lithium deposit, or a coastline. The same essay argues that the data-center buildout itself converts abundant intelligence straight back into demand for power, water, and configured matter — and that at the far end, genuinely cosmic projects would be bottlenecked on energy and compute allocation, not on ideas. Bits are cheap; atoms are not.
Status, attention, and the positional trap
Some goods are scarce by definition: their worth comes from others not having them. Status, rank, and being first cannot be mass-produced, because abundance is precisely what they exclude. Human attention is the same — there is a fixed daily supply of it per person, and no model makes a day longer. As output floods every channel, the scarce thing is not another piece of content but a reason for a human to look, and the standing that confers it.
Authenticity, trust, and provenance
When anything can be generated convincingly, the premium lands on provenance — knowing a thing is what it claims to be. A handmade object, a live performance, a verifiably human judgment, an emotion that is actually felt: these draw value from a backstory that cannot be synthesized after the fact. Trust becomes a scarce asset because counterfeiting everything else gets cheap. The signature that matters is no longer skill, which machines supply in bulk, but authenticated origin.
The directional goods
The article's sharpest claim is that the rarest commodity becomes direction — the authority and wisdom to decide what powerful systems are pointed at. It frames stewardship, the setting of values and constraints, and accountability for outcomes as the enduring human work, "orders of magnitude larger" than the cognitive labor being automated. This is a positional and political good as much as a moral one: influence over civilizational steering is, by nature, concentrated and contested.
A pattern worth holding loosely
The relocation thesis is a useful lens, not a forecast. Whether intelligence actually goes "too cheap to meter," and on what timeline, is genuinely uncertain — the underlying analysis hangs on adoption curves and energy limits that could bend either way. But the structural point survives the uncertainty: a world that makes minds cheap does not become a world without scarcity. It becomes a world where what is scarce is harder to fake, harder to scale, and more human.
Related concepts
Appears in
Details
- Section:
- Notes
- Updated:
- 2026-06-26