For most of its history, economics has rested on a quiet but foundational assumption: human cognition is scarce.
Thinking is slow. Judgment is costly. Expertise accumulates over years. Decision-making capacity is bounded. Every major economic model—implicitly or explicitly—treats cognition as an expensive input that constrains production, coordination, and allocation.
Artificial intelligence breaks this assumption.
Not gradually. Not marginally. But categorically.
When cognition becomes cheap, economics does not merely need new parameters. It needs new primitives.
Classical and neoclassical economics rarely model cognition directly, yet it sits everywhere in the background.
Labor is productive because it embeds skill and judgment. Firms exist because coordination is cognitively expensive. Wages reflect productivity because output is tied to human decision-making. Markets fail when information is asymmetric or costly to process.
Even modern growth theory, behavioral economics, and institutional economics all retain the same structural premise: thinking is a limiting factor.
Artificial intelligence removes that limit.
Once judgment, analysis, synthesis, and even creativity can be produced at near-zero marginal cost, the scaffolding that holds many economic intuitions in place begins to wobble.
Most technological progress improves tools. AI changes inputs.
Steam amplified muscle. Electricity extended energy. Computers accelerated calculation. AI substitutes for cognition itself.
An economy can absorb faster tools without revising its logic. It cannot absorb abundant cognition without altering how value is created, distributed, and defended.
When a resource moves from scarce to abundant, it stops being priced and starts being assumed. We no longer price air. We no longer allocate daylight. And increasingly, we will no longer allocate thinking.
This is not a productivity curve. It is a regime shift.
For centuries, effort was a reasonable proxy for value creation: more hours, more skill, more thinking—more output. AI breaks that coupling.
A single human, augmented by cheap cognition, can now produce output that once required teams, departments, or entire firms. Conversely, many forms of effort become economically redundant regardless of how skilled or diligent the worker is.
This creates a paradox for labor economics: effort remains psychologically meaningful but economically optional.
Wages, which evolved to price scarce human contribution, lose their anchoring mechanism when contribution no longer requires human cognition at scale.
It would be a mistake to conclude that markets fail in a post-scarcity cognition environment. Markets do not disappear when inputs change; they reorient.
When cognition is abundant, markets stop optimizing for production efficiency and begin optimizing for:
• Ownership of capital and infrastructure
• Control of distribution
• Access to attention
• Legal and institutional leverage
• Speed of deployment, not quality of thought
In other words, economic power migrates upstream.
Value accrues not to those who think best, but to those who own the systems that think.
Ronald Coase explained firms as a response to transaction costs. AI collapses many of those costs.
Coordination becomes cheaper. Planning becomes faster. Monitoring becomes automated. Decision-making becomes scalable.
This does not eliminate firms—but it reshapes them.
Firms become less about organizing labor and more about:
• Owning models
• Owning data
• Owning interfaces
• Owning legal and regulatory moats
The firm survives, but labor retreats from its center.
When cognition is cheap, income cannot be justified purely by contribution.
This destabilizes long-standing moral narratives embedded in economic systems:
• Wages reward effort
• Skill acquisition guarantees security
• Productivity growth benefits labor broadly
None of these claims survive intact.
Redistribution debates, universal basic income, and social safety nets are often framed as ideological conflicts. In reality, they are accounting responses to a system where labor income is no longer the primary distribution mechanism.
The politics will be loud. The cause is mechanical.
Because it is uncomfortable.
It undermines:
• Labor-centric growth models
• Human capital theory as a universal solution
• Meritocratic narratives embedded in policy
• Institutional designs built around employment as a proxy for participation
Most importantly, it forces economics to admit that intelligence was doing more work than theory acknowledged. AI makes that visible.
The unresolved question is not whether cognition will become cheap. It already is.
The question is whether economic theory will treat this as:
• A temporary distortion to be absorbed, or
• A structural event requiring new foundations
History suggests that when inputs change category—from scarce to abundant—theory follows slowly, often after institutions begin to fail quietly.
This essay does not propose a new economic model. It names the rupture. Others will fill in the math.
@article{Asefi2025CognitionCheap,
title = {When Cognition Becomes Cheap: The Event Economics Is Avoiding},
author = {Houman Asefi},
year = {2025},
url = {https://houmanasefi.co/essays/when-cognition-becomes-cheap-the-event-economics-is-avoiding.html}
}