Cognitive Mercantilism
Essay by Venkatesh Rao on Contraptions (2026-10-25) arguing that current AI panic is a form of cognitive mercantilism — the same category error that pre-capitalist economic thought made about material wealth, applied now to intelligence. Reading Joyce Appleby's The Relentless Revolution: A History of Capitalism (2010), Rao draws a four-point structural parallel and extends it to a positive thesis: a Cognitive Industrious Revolution in which abundant intelligence stimulates new voluntary cognitive effort the way imported sugar and cottons stimulated the industrious revolution of early modern Britain.
The mercantilist worldview, mapped to intelligence
Rao's central move: swap "wealth" for "intelligence" in the four pre-capitalist assumptions and the parallels become "embarrassingly direct."
| Mercantilism (wealth) | Cognitive mercantilism (intelligence) |
|---|---|
| Wealth is scarce and finite; hoard bullion, favorable balance of trade | Intelligence is fixed, embodied in brains and IP; hoard talent, patents, model weights |
| Ordinary people cannot be trusted; govern coercively "for their own good" | Ordinary people cannot be trusted with intelligence; cognitive elites must regulate who can use it |
| Economy is supply-side; demand barely registers; mass as factor of production | AI discourse is supply-side; demand barely registers; mass as consumer of elite-chosen outputs |
| Most of humanity is good only for subsistence farming — remove them and you get "useless mouths" | Most of humanity is good only for low-grade cognitive labor — automate it and you get "useless minds" |
quote "The cognitive mercantilist wants to hoard intelligence and its products like bullion. The global competition for talent becomes a treasure hunt for genius brains and the valuable knowledge species they can produce, which now includes AIs in addition to copyrightable culture and patentable inventions."
The industrious revolution — the historical hinge Rao uses
Appleby's book argues capitalism was not inevitable but a historically contingent invention that required escaping zero-sum assumptions. The hinge wasn't factories; it was demand. Before the Industrial Revolution, ordinary Britons developed tastes for tobacco, tea, sugar and Indian cottons — goods that were neither subsistence nor aristocratic markers — and voluntarily increased annual working days from ~250 to ~300 to afford them. The surplus income sustained demand, demand incentivized productivity, productivity generated more surplus. "A virtuous cycle was taking shape." This happened a century before organized labor politics.
Rao reads modern labor politics as partly this: the politics of relative deprivation in a society of expanding possibilities. "Not merely freedom from coercion, but freedom to participate in the expanding possibilities of economic life." He cites C.K. Prahalad's "fortune at the bottom of the pyramid," and the Western-do-gooder blind spot that people in slums will cut food spending to buy a TV.
The Cognitive Industrious Revolution
Rao's positive thesis — see dedicated page Cognitive Industrious Revolution.
quote "The crucial economic question then becomes not how much intelligence people can produce, but how much intelligence they can find uses for. And there is no obvious upper bound on that demand."
Key mechanism: many people who appear stupid are not incapable of thought — they are not motivated to think given poor returns on cognitive effort. Abundant AI changes that calculation. A modest investment of attention unlocks previously-inaccessible projects, so the returns on attention rise, and with them the incentive to pay attention. "Someone who previously had little reason to think hard about a problem may suddenly discover that a modest investment of attention unlocks enormous possibilities."
Rao reports this of himself: "I'm suddenly willing to work significantly harder compared to even a year ago, because a lot more is suddenly within reach."
The analogy's teeth: abundant food did not kill cuisine; it enabled elaborate culinary cultures. Abundant intelligence may not kill thinking; it may enable more ambitious cultures of thought — because people no longer have to do all the cognitive production themselves.
From mercantilism to Malthusianism
See AI Malthusianism. Rao distinguishes two adjacent failures of imagination:
- Cognitive mercantilism fears losing control of scarce intelligence.
- AI Malthusianism fears the consequences of abundant intelligence — that surplus intelligence will outrun the fixed capacity of human minds to absorb it, and "turn toxic in some way."
Both assume the demand for intelligence is fundamentally limited. Both miss Buckminster Fuller's ephemeralization (doing more with less), the way the agricultural revolution did not create useless mouths but freed people to want new things.
The demand side of AI — the radical extension
Rao's most speculative move: why should AI systems be regarded exclusively as producers of intelligence rather than potential consumers of it?
quote "What if they could be unlocked as new sources of aggregate demand, as happened with subsistence farmers? ... An AI agent pursuing open-ended projects, discovering new interests, collaborating with other agents, and generating its own reasons to acquire further intelligence would be participating in a cognitive economy as something more than a factor of production or paperclip-maximizing externality."
He frames the market itself as the original emergent planetary intelligence, where many products exist because the market demands them, with humans as secondary consumers. From this angle, contemporary AI safety anxieties begin to resemble early modern elite fears of emancipated agricultural laborers — "won't they turn monarchs and nobles into paperclips?" Freedom creates real risks, but the risk of autonomy is not an argument for permanent servitude.
Labor politics of cognitive abundance
Rao's forecast: the central political demand will not be preservation of existing cognitive jobs (any more than industrial labor demanded preservation of agricultural employment). It will be a fair share of expanding cognitive consumption opportunities.
quote "People may cut back on streaming subscriptions in order to be able to afford a coding agent, just as they once cut back on groceries to buy TVs. It is perhaps not a coincidence that both Claude Code and Netflix Standard cost around $20/month today."
Access to intelligence, the ability to command it, the freedom to pursue projects with it, and the institutional rights to participate in an intelligence-rich economy — these become the foundations of a new labor politics. Cognitive abundance, not cognitive scarcity.
Why this matters for the brain
Rao's piece is a framing contribution, not a numerical one — but it reframes several threads in the vault:
- The Capability Gap concept gains a demand-side companion: the ceiling isn't only "the ceiling is rocketing"; it is "abundant intelligence creates new demand that expands what people want to do." The failure mode under Rao isn't insufficient rate of change — it's not finding uses for the intelligence newly within reach.
- The Tasks to Responsibilities Shift gets re-cast: delegating tasks to agents frees humans to be consumers of intelligence pursuing projects, not merely orchestrators of cognitive serfs.
- The Solopreneur AI Economy slots in cleanly — one person commanding quantities of intelligence previously requiring teams is exactly Rao's "I can enjoy food without knowing how to farm" analogy.
- The Upskilling Bill gains a demand-side rationale: upskilling as pursuit of consumption opportunities, not merely defensive adaptation to displacement.
- DRAG for AI Upskilling at Manila IT Site and Unlocking 10X in Domain Masters as AI Gets Better both read as implementations of Rao's thesis at the enterprise scale: raise individual returns-to-attention and voluntary cognitive effort rises.
Open threads
- Rao makes no empirical claim. The thesis predicts a cognitive industrious revolution; evidence of one would be rising voluntary cognitive effort tied to AI access, not merely rising adoption. The vault's closest proxies: Ramp/Revelio adopter headcount growth (see AI Adoption and Headcount Growth), Naukri +25%/-3% split (in Unlocking 10X in Domain Masters as AI Gets Better), Faros 98% more PRs / 91% more review time (ambiguous — more work done, more work reviewed; who is more ambitious?).
- Rao's AI-as-consumer-of-intelligence thesis is speculative and will stay so until there is a plausible economic account of autonomous AI demand that isn't just "the operator pays." Worth filing a
querypage if the discourse develops. - The $20/month Claude Code vs Netflix Standard parallel is a striking rhetorical hook; test: does the retention + spend pattern of consumer AI subscriptions actually resemble streaming or something new?
Secondary coverage
- 2026-11-02 — Nathaniel Whittemore devoted ~43 min of The AI Daily Brief to the essay (episode: "AI Won't Make Our Minds Useless According to 400 Years of History"). Fast multi-channel pickup — ~8 days from Rao's publication to a dedicated AIDB episode. First captured blurb-only in AI Daily Brief Episode Snapshot (Late October 2026); full transcript ingested 2026-11-03 as AI Wont Make Our Minds Useless According to 400 Years of History (AI Daily Brief). Whittemore adds two frames the essay lacks, Six Elasticities of AI Demand (demand expands where a named constraint is removed, not on its own) and Infinite Backlog (the anxiety-side twin of Rao's "I work harder now"), plus a caveat Rao does not make: optimism about the destination is compatible with concern about the transition. He also surfaces Rao's "mercantilism was not for merchants" section and the 80%→95% line, both under-covered above.