Notes
The Economics of Abundance
On this page
The economics of abundance is what happens to prices, work, and buying power when the marginal cost of cognition falls toward zero. For most of history, thinking was scarce: every diagnosis, contract, lesson, or line of code required a trained human and their time. As AI systems begin to supply judgment-shaped work at near-zero incremental cost, that scarcity dissolves — and the price system built around it starts to behave strangely. Some things get radically cheaper while others get dramatically more expensive, often at the same time.
A two-tier price regime
The central pattern is not simple deflation but a split. Anything that can be replicated by an agent — software, digital media, routine analysis, first-draft work — drifts toward its falling production cost. Early signs are already visible: disinflation in software and content, and firms reporting large gains in revenue per worker. Goldman Sachs has estimated that roughly 300 million full-time-equivalent jobs are globally exposed to automation, and BCG modeling suggests half or more of US jobs could be reshaped within a few years.
Meanwhile the things AI cannot cheaply copy move the other way. Housing in productive cities, energy for data centers, live human attention, and skilled oversight of AI systems hold or gain value. The result is a two-tier economy where your tools keep getting cheaper while the things that signal status, trust, or genuine presence keep getting dearer.
The phases of the transition
The shift is best read as a sequence rather than an event.
In the agentic present, agents compress the cost of discrete tasks. Capital floods AI infrastructure faster than labor markets adjust, producing wage premiums for AI fluency alongside hiring slowdowns in exposed roles — a classic case of capital capturing productivity gains before labor catches up.
In the intelligence explosion that follows, generality and then recursive self-improvement push deflation deep into physical and cognitive goods at once. Energy and compute become the visibly strategic resources — "the new oil." The strain here is structural: as more value is produced without wages, the wage-based mechanism that historically distributed purchasing power begins to break down. Abundance on the supply side can collide with a demand-side crisis if too few people have income to buy what the machines make.
In the post-singularity economy, material scarcity is largely solved for those with access. "Money" increasingly names coordination tools for a new scarce layer — compute and expansion rights, reputation ledgers, and "meaning dividends" funded directly by machine productivity rather than labor.
What never gets cheap
Abundance is not uniform, and the article is careful about its limits. What resists the collapse is a short, stubborn list: high-fidelity human connection, the ability to keep superintelligent systems pointed at human ends (alignment capacity), energy for the largest projects, and a felt sense of meaning and contribution. These become the genuinely scarce goods — and the new sources of inequality, since access to augmentation diverges sharply between an augmented core and everyone else.
The honest reading is therefore mixed rather than utopian. The marginal cost of intelligence falling toward zero is real, and so are the transition strains: capital-labor divergence, a possible demand collapse if distribution is not redesigned, and a divide between the augmented and the unaugmented. The work of intention, judgment, and stewardship does not disappear in this picture. It becomes the scarce thing — and, the argument runs, more visible and more valuable than it has ever been.
Related concepts
Appears in
The Economics of Abundance: AI, Transformation, and the Singularity
Details
- Section:
- Notes
- Updated:
- 2026-06-26