Value Creation vs. Value Extraction

In the AI Story

The distinction emerged from Mazzucato's 2018 book The Value of Everything, which traced the history of how economic theory progressively lost the capacity to distinguish productive from unproductive activity. Classical economists from Adam Smith through Marx made the distinction central to their frameworks. Twentieth-century neoclassical economics abandoned it, treating all market transactions as presumptively value-creating so long as willing parties transacted at market-clearing prices.

Applied to the AI economy, the distinction illuminates dynamics the innovation narrative routinely conflates. An AI tool that enables a person who previously could not create software to build something genuinely useful is creating value — the person's capability is expanded, the application serves a previously unmet need, human welfare improves. An advertising optimization algorithm that uses AI to capture more consumer attention and direct more purchasing decisions is doing something structurally different — it does not create the consumer's need, it captures existing purchasing power by exploiting informational asymmetries.

The confusion between creation and extraction is systematically maintained by the way the AI economy presents itself. Financial metrics — revenue, profit, market capitalization — measure the capture of value, not its creation. A company that extracts a billion dollars through advertising optimization may report higher profits than a company that creates genuine value through AI-assisted healthcare. The conventional wisdom does not distinguish between the two, and the market rewards the former as enthusiastically as the latter.

The distinction has a temporal dimension. Value creation tends to produce returns distributed over time — the builder who creates a useful application generates value for its users throughout the application's lifespan. Value extraction tends to concentrate returns in short bursts — the algorithm that exploits a pricing inefficiency captures value in a single transaction. Patient value creation builds lasting economic capacity; rapid value extraction depletes it.

Origin

Mazzucato developed the distinction through her analysis of financial-sector returns following the 2008 crisis, when she argued that much of what finance describes as value creation is structurally value extraction — returns captured from productive sectors without proportionate contribution. Her 2018 book The Value of Everything: Making and Taking in the Global Economy systematized the framework and traced its intellectual history from Aristotle through the Physiocrats, classical political economy, and the neoclassical turn that erased the distinction.

Her application to AI, developed across essays, interviews, and the Algorithmic Rents research program, extends the framework to the specific extractive dynamics of platform economies. The 2024 Guardian commentary and the February 2025 Project Syndicate essay Governing AI in the Public Interest made the distinction central to her AI policy prescriptions.

Key Ideas

Functional, not organizational. Both public and private entities can create or extract value; the test is what they produce, not who owns them.

Market metrics are insufficient. Revenue and profit do not distinguish creation from extraction — both can produce identical financial statements.

Temporal asymmetry. Creation distributes returns over time; extraction concentrates them in bursts. This shapes the sustainability of economic activity.

Amplifier neutrality. AI as technology amplifies both creation and extraction with equal efficiency — the institutional architecture determines which dominates.

Policy implications. Tax policy, competition law, patent frameworks, and labor policy should treat the two activities differently — rewarding creation, constraining extraction.

Debates & Critiques

Critics argue the distinction is subjective — what looks like extraction to one observer may be creation to another, and economists have long since abandoned the search for objective value measures. Mazzucato's response is that refusing to make the distinction does not eliminate the underlying reality; it simply allows extraction to proceed under the cover of rhetoric about innovation and disruption. The empirical evidence — profit margins that exceed any plausible return on genuine innovation, concentrations of market power that no competitive market produces — supports the analytical distinction whether or not mainstream theory acknowledges it.