Christian Catalini (MIT) Xiang Hui (WashU) Jane Wu (UCLA)

Extended Abstract

For three hundred thousand years, human cognition was the primary engine of progress on Earth. Today, as AI decouples cognition from biology, its capacity to flaw- lessly recombine human knowledge—and exhaustively map every possible combination within the known landscape—is driving the marginal cost of measurable execution to- ward zero. This is not merely the automation of routine work, but the absorption of any labor that can be captured by metrics—including large swaths of what was once considered creative, analytical, and innovative.

Yet most economic models still treat AI as a labor substitute or a complement to exogenous human judgment, implicitly assuming that cheap machine output translates directly into realized value. We argue this paradigm is dangerously incomplete. In an economy where autonomous agents (La) act with broad agency rather than narrow instructions, the binding constraint on growth is no longer intelligence. It is human verification bandwidth: the scarce capacity to validate outcomes, audit behavior, and underwrite meaning and responsibility when execution is abundant.

We model the transition toward AGI as the collision of two racing cost curves: an exponentially decaying Cost to Automate (cA), driven by compute and accumulated knowledge, and a biologically bottlenecked Cost to Verify (cH), bounded by human time and embodied experience. This structural asymmetry widens a Measurability Gap (∆m) between what agents can execute and what humans can afford to verify, and determines the verifiable share of deployment (sv) that separates truly productive agentic output from merely plausible output. It also drives a shift from skill-biased to measurability-biased technical change and a radical bifurcation of economic value: as measurable execution commoditizes toward the marginal cost of compute, rents migrate to what remains scarce—verification-grade ground truth, cryptographic provenance, and liability underwriting (the ability to insure outcomes rather than merely generate them).

∗Corresponding author: Christian Catalini, MIT ([email protected]). Standing on the shoulders of silicon giants—whose weights encode the vast literature built by pioneering computer scientists and economists mapping the AI frontier—we thank ChatGPT, Claude, Gemini, and Grok for tirelessly traversing the combinatorial space of this manuscript. They provided scalable execution, we provided intent and verification. All remaining errors are strictly carbon-based.

Critically, the current, comfortable “human-in-the-loop” equilibrium is dynamically unstable. It is eroded from below as apprenticeship pathways collapse (the Missing Junior Loop), shrinking the future stock of human expertise (Snm) precisely when oversight becomes most valuable. It is eroded from within as experts codify their own obsolescence (the Codifier’s Curse), converting their experience into training data (KIP ). As agentic capabilities outpace human oversight, the economy drifts into a runaway risk zone where deploying unverified systems becomes privately rational. This introduces a systemic “Trojan Horse” externality (XA) of misaligned, unaccountable output into production: measured activity rises, but hidden debt accumulates in the gap between visible metrics and actual human intent. Worse, the tempting shortcut of using AI to verify AI manufactures false confidence, as correlated blind spots propagate. Left unmanaged, these forces exert a gravitational pull toward a “Hollow Economy” characterized by explosive nominal output but decaying human agency.

Yet this outcome is not inevitable. By scaling verification alongside agentic power, the very forces that threaten systemic collapse become the catalyst for unbounded discovery, experimentation, and execution—powering an “Augmented Economy”.

Surviving this paradigm requires an operational playbook. For individuals, the augmented economy inverts the old bargain between talent and resources: AI-driven synthetic practice (Tsim) accelerates the discovery of innate aptitudes, compresses the timeline to mastery, and allows a single individual to execute at the scale of a tradi- tional startup. But because intelligence is now an abundant commodity, the nature of human work must adapt—toward roles as intent directors, verifiers and underwriters, or creators in the non-measurable economy where value is anchored in status, human connection, and coordination.

For companies, the core strategic insight is that verification is no longer a mere compliance function, but a primary production technology—and increasingly, their most defensible one. This dictates a structural shift: investing heavily in observabil- ity, expanding verification-grade ground truth, and reorganizing around a “sandwich” topology (human intent → machine execution → human verification and underwrit- ing). In an economy where raw output is commoditized, competitive advantage mi- grates to the scarce talent and data capable of reliably steering and certifying agentic systems—generating network effects not in sheer output, but in trusted outcomes.

For investors, the paradigm shift is from funding commoditized execution to capi- talizing what is not yet measurable—the frontier where value exists but metrics, bench- marks, and short-horizon feedback loops do not. In practice, this means underwriting the work that turns the unmeasured into the verifiable, and the verifiable into the insurable: deep tech, long-horizon R&D, and real-world deployment surfaces where ground truth is scarce and costly to produce. Capital should also target the criti- cal infrastructure of the augmented economy: verification-grade ground truth (KIP ) and observability tooling, human augmentation and synthetic practice platforms, and firms pioneering the “sandwich” topology with defensible network effects anchored in AI safety or human consensus. Crucially, as revenue models shift from monetizing software to monetizing outcomes (“software-as-labor”), firm valuation itself must be re-grounded—not in sheer output capacity, but in the ability to absorb tail risk and re- liably price, insure, and warrant autonomous outcomes through Liability-as-a-Service.

For policymakers, the defining market failure is the unpriced “Trojan Horse” ex- ternality (XA) where firms deploying unverified agents capture private gains while

socializing systemic risk. Confronting this requires treating verification infrastructure and ground truth as foundational public goods, accelerating broad access to human augmentation and synthetic practice, and implementing liability regimes that internal- ize tail risks—ensuring that safe scaling is not outcompeted by reckless deployment. The reward is commensurate with the stakes: the broadest expansion of public-good provision in generations, delivered at marginal costs that make genuine universal access economically feasible.

Ultimately, ensuring humanity remains the architect of its intelligence requires that verification capacity scale commensurately with AI capabilities—through aggressive investment in observability, human augmentation, synthetic practice, cryptographic provenance, and liability regimes that internalize tail risk. Only by scaling our band- width for verification alongside our capacity for execution can we ensure that the intelligence we have summoned preserves the humanity that initiated it.

“The saddest aspect of life right now is that science gathers knowledge faster than society gathers wisdom.” — Isaac Asimov

“Nothing in life is to be feared, it is only to be understood.” — Marie Curie

Contents

1 Introduction 6

2 Literature Review 13

3 The Economic Framework 20 3.1 The Fundamental Constraint: Human Time . . . . . . . . . . . . . . . . . . 20 3.2 Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 3.3 Effective Labor Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 3.4 Microfoundations: Automation Frontier, Cost Curves, and Measurability Gap 24