Some Simple Economics of AGI

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 flawlessly recombine human knowledge—and exhaustively map every possible combination within the known landscape—is driving the marginal cost of measurable execution toward 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).

Introduction. For three hundred thousand years, human cognition was the main engine of progress on Earth. Yet this profound intelligence is highly idiosyncratic. Forged in the unforgiving environment of natural selection, our minds are an exquisite artifact of biology—an “animal intelligence” fine-tuned by the harsh reward functions of metabolic efficiency, physical survival, and social friction. Every monument of human civilization, from Euclidean geometry to the Apollo program, emerged from this specific, biologically constrained architecture. Today, we are decoupling intelligence from biology. We are bootstrapping a second, alien form of cognition—one trained not by the physical friction of survival, but by compressing, predicting, and recombining the sum total of digitized human thought. This new intelligence does not inherit our muscles, our hormones, or our evolutionary priors. It inherits something else: a vast latent map of what our species has written, drawn, coded, and measured. In doing so, it forces an ancient question into a modern frame: asking not merely what we can build, but fundamentally what we are. As this synthetic intelligence comes online, the semantic debate over whether it constitutes “Artificial General Intelligence” (AGI) is rapidly becoming a distraction. The exact threshold of human parity is economically and philosophically moot, because the specialization of these intelligences is fundamentally different. Human cognition evolved for survival under severe biological constraints; machine cognition is built for scalable search and simulation over vast, learned representations. Over the next 12 to 24 months, the global economy will be transformed not by a machine that is “exactly like a human,” but by systems that can traverse the combinatorial space of human knowledge with a fidelity and scale that biology simply cannot match. If innovation is often the cross-pollination of ideas across disparate domains (Jacobs, 1961), then an entity capable of holding the entirety of recorded thought in its latent space will unlock unprecedented value through synthesis alone—surfacing connections no individual, organization, or discipline has the time, attention, or institutional memory to notice. The deeper acceleration, however, stems from the fact that this map will not remain frozen. Today’s foundation models inherit latent maps fixed at training time: rich historical snapshots that do not update as the world unfolds. The rapid convergence of real-time retrieval and continual learning will steadily dissolve that constraint. Retrieval allows these systems to dynamically ingest the unfolding present, while continual learning allows them to retain what they learn—progressively integrating interactions, feedback, and real-world outcomes into their priors. By closing the loop between observation and learning, they transition from static archives into dynamic agents. They will no longer merely map the known; they will instrument it, test it, and extend it—probing the boundary where the map fails and turning portions of the unknown into new terrain for search. In that regime, recombination ceases to be a human-gated batch process over a frozen snapshot and becomes a continuous, compounding loop: observe →retrieve →act →learn. The frontier will advance simply because recombination approaches a marginal cost of zero. Yet, it is precisely this frictionless acceleration that will soon overwhelm our capacity to verify what the machine has discovered or built. This shift implies a radically new form of disruption. Historically, automation was bounded by routine tasks and explicit codification. In the agentic economy, however, the dividing line is no longer routine versus non-routine, but measurable versus non-measurable. Any task that can be reduced to a metric can be industrialized (Catalini et al., 2025)—regardless of the prestige or complexity previously associated with its human execution. More profoundly, AI acts not merely as a labor substitute within the already-measured world; it is an engine for expanding measurement itself. By converting previously unstructured phenomena into structured data streams, it continuously enlarges the domain of what can be automated. Until recently, it was comforting to assume that synthetic intelligence was strictly bounded by what humanity had already mapped and measured. We retained a monopoly on the “unknown unknowns”—the domains of Knightian uncertainty where humans step off the map, relying on embodied experience and hard-won intuition. However, as these systems evolve from pattern matchers into dynamic agents that plan, experiment, and simulate—culminating in Large World Models capable of representing physics, causality, and counterfactuals—they threaten to shatter this final constraint. By exhaustively simulating counterfactual realities, agents will increasingly internalize the capacity to deduce optimal policies in entirely novel environments.

Related work. This literature review situates the paper within multiple research streams: (1) task-based models of automation and labor markets, (2) the emerging empirical evidence on LLM adoption, productivity effects, and AI exposure, (3) principal–agent theory and its application to AI alignment, (4) the microstructure of human capital formation, training, and learningby-doing, (5) technical AI safety research on distributional shift, reward hacking, learned optimization, and model collapse, and (6) the economics of digital markets, platform competition, and value in an age of abundance. Establishing a definitive, visionary roadmap for the economics of AI, Brynjolfsson et al. (2025b) emphasize that modeling the macroeconomic trajectory of transformative AI requires dismantling the artificial boundaries between micro-level labor disruptions, growth constraints, and technical alignment. Rather than treating these as isolated domains, this review brings them together to illustrate how the shifting boundary between agentic execution and human verification reshapes labor dynamics, digital markets, and organizational risk. The dominant paradigm for analyzing the labor market impact of automation is the task-based model pioneered by Autor et al. (2003) and formalized into a general equilibrium framework by Acemoglu and Restrepo (2018). In this approach, jobs are decomposed into constituent tasks, some of which are susceptible to automation while others remain the province of human labor. Automation displaces workers from specific tasks but simultaneously generates reinstatement effects as new, complementary tasks emerge for which human labor holds a comparative advantage. Autor et al. (2003) establish the foundational distinction between routine tasks—susceptible to automation because they follow explicit, codifiable rules—and non-routine tasks. Acemoglu and Restrepo (2018) formalize this into a growth model, characterizing the “race between man and machine” as a dynamic tension between the displacement effect and the reinstatement effect. Whether automation raises or lowers aggregate wages depends on the relative pace of these two forces. This paper builds directly on the insight of Autor (2015) that the boundary of automation is a moving frontier. However, it argues that the critical delineation is no longer routine versus non-routine, but measurable versus non-measurable. The measurability gap ∆m = mA −mH formalizes Polanyi’s paradox (Polanyi, 1966) applied to verification: tasks become automatable not when they are simple, but when they can be measured (Catalini et al., 2025), creating a risk zone where execution is cheap but verification remains tacit and expensive.

Method. 3 The Economic Framework The model builds on Romer’s seminal endogenous growth theory (Romer, 1990), but introduces a few, key distinctions. First, while standard endogenous growth models microfound technical change via the horizontal expansion of intermediate capital varieties, our framework reorients the locus of innovation around the vertical automation of tasks. This allows us to focus theoretical tractability on the composition of effective labor, abandoning traditional labor aggregates—such as the skilled-versus-unskilled dichotomy—to decompose work strictly along the axis of measurability. We model Effective Labor as requiring a complementary balance between Measurable Capacity (Lm), where scalable agentic labor acts as a perfect rival and substitute for human execution, and Non-Measurable Capacity (Lnm), which provides the scarce, biologically bottlenecked human steering and verification required to direct that execution. Second, we abandon the traditional assumption that automated capability unconditionally translates into productive economic value. In our model, agentic labor only contributes to effective capacity when filtered through an endogenous verifiable share (sv). The unverified residual leaks into the economy as a “Trojan Horse” externality (XA) that actively parasitizes the general capital stock and crowds out human consumption. Last, rather than treating human expertise as a static endowment or a naturally accumulating byproduct of education, we formalize the dynamic fragility of human capital. Because the non-measurable expertise (Snm) required to safely oversee AI is built through the friction of routine execution, automating measurable work actively destroys the training ground for future verifiers. This introduces a profound structural asymmetry: the economy’s capacity to scale execution expands precisely as its biological capacity to verify and steer that execution decays. To unpack the source of this verification bottleneck, our framework begins by formalizing the underlying constraint that limits this biological capacity: the finite budget of human time.

3.1 The Fundamental Constraint: Human Time • Tm (Measurable Work): Time spent on tasks occurring in environments that can be extensively measured and codified, transforming reality into data streams for automation. This extends beyond routine execution to include any complex domains where a phenomenon can be shoehorned into numbers (Catalini et al., 2025).

• Tnm (Non-Measurable Work): Time spent on tasks that cannot be successfully measured yet—often because measurement is currently thwarted by extreme difficulty, prohibitive costs, or privacy and regulatory barriers. This includes domains characterized by Knightian uncertainty, where outcome probabilities are fundamentally unknowable.

We build on the formalization of human judgment developed by Agrawal et al. (2018) and Agrawal et al. (2025a) by anticipating a substantially stronger, more autonomous role for artificial intelligence. Agrawal et al. model judgment as a set of human cognitive capacities: the ability to recognize opportunities for improvement (opportunity judgment) and to map uncertain states to optimal actions (payoff judgment).

However, as AI agents achieve continual learning, they increasingly internalize these exact capacities. By exhaustively exploring recombinations within known parameter spaces, agents can already autonomously identify opportunities; by leveraging world models to simulate scenarios and counterfactuals, they will soon be able to deduce optimal state–action mappings.

Whereas their framework centers on cognitive limits in tool-assisted production— including the ability to map states to optimal actions and recognize high-quality output (“payoff judgment”)—ours shifts the friction to contract theory and the economics of delegation. Even a principal who can recognize good output on inspection may remain unable to contract on that quality—because the criteria that make it good are tacit, the evaluation is context-dependent, and the feedback loop cannot be codified into a metric that scales without the principal’s direct oversight.

Discussion. 7 Strategic Implications The model describes an economy undergoing a structural phase transition toward the post- AGI frontier. This section summarizes our core insights, which we then translate into actionable strategies for key stakeholders in the next.

7.1 The Era of Measurability-Biased Technical Change The model focuses on a novel form of disruption: measurability-biased technical change. Historically, economists focused on “skill-biased” technical change—a regime where education protected wages and automation was strictly limited to repetitive, explicitly codifiable actions. In our framework, however, jobs are bundles of tasks distinguished not by skill, but by measurability. Tasks that can be measured (Tm) can be automated, regardless of their complexity or the educational credentials required to perform them (Catalini et al., 2025). This insight creates the central existential challenge of the agentic economy: automation no longer targets only the routine execution characteristic of previous industrial waves; it targets anything we can collect data for. This dynamic is exacerbated by the fact that AI itself is a powerful engine of measurement—converting previously unstructured phenomena into numbers (e.g., through computer vision, digitized interactions, agentic feedback loops, etc.)—thereby actively expanding the automation frontier.

7.2 Verification as a Key Bottleneck to Safe Deployment Agents only contribute fully valid economic capacity if their output can be verified. This constraint explains why early product-market fit clustered around chatbots and image generation: domains where human verification is intuitive and the cost (cH) is negligible relative to the value generated. However, as agents take on longer and more complex tasks, the verification dynamic shifts. The human capacity to cost-effectively audit intent and validate edge cases becomes a key constraint. Crucially, this bottleneck does not stop deployment: instead, the relentless economic pressure to automate (cA →0) incentivizes unverified deployment—a regime where we implicitly bypass the human loop and allow hidden risks (XA) to accumulate rather than slowing down for human verification. In the model, this transition is driven by the divergence of two cost curves:

• Cost to Automate (cA): Driven down exponentially by compute (KC) and accumulated knowledge (public A and proprietary KIP).

• Cost to Verify (cH): Driven up by feedback latency (tfb) and the scarcity premium of expertise (w(Snm)). The latter can create a situation where for some tasks verification becomes economically unviable even if technically possible.

As cA collapses faster than cH, the measurability gap (∆m) widens: we become increasingly capable of generating output that we are decreasingly capable of verifying.

7.2.1 The Structural Instability of the “Human-in-the-Loop” The “human-in-the-loop” equilibrium is inherently fragile due to a fundamental mismatch in scaling laws: while agentic labor scales elastically with compute and data, human verification remains inelastic—bound by the slow, biological accumulation of embodied experience (Snm). This mismatch manifests as four structural failures that systematically erode the conditions for safe verification:

1. The Missing Junior Loop: by automating entry-level measurable work (Tm), we destroy the training ground required to build future experts (Snm). Without involvement in execution, the pipeline of senior verifiers with the intuition to properly check the agents dries up.

2. The Codifier’s Curse: as experts apply their Snm to verify tasks, they inevitably generate labels that digitize their tacit intuition into new KIP. In doing so, they fuel the very automation that displaces them.