When artificial general intelligence (AGI) reaches the point where it can perform the cognitive work humans do, robots will proliferate rapidly, and the economic consequences could be staggering. According to analysis of standard US industrial data, a fully automated economy could double its output roughly every year, with no additional technological breakthroughs required. This scenario represents one of the most consequential economic shifts in human history, yet policymakers and business leaders are only beginning to grapple with its implications.

What Would an Automated Economy Actually Look Like?

The concept of AGI triggering an "industrial explosion" rests on a straightforward economic principle. Once AI systems can replicate human cognitive abilities, the marginal cost of labor approaches zero for any task that can be digitized or performed by robots. This creates a self-reinforcing cycle: as automation reduces production costs, companies can manufacture more goods and services. Those goods and services, in turn, can be used to build more robots and manufacturing capacity. The result is exponential growth in physical output.

This isn't speculation about distant futures. The economic models underpinning this projection rely on input-output analysis, a well-established framework that economists have used for decades to understand how industries depend on one another. When you feed realistic automation assumptions into these models using actual US industrial data, the numbers suggest that doubling output annually is mathematically plausible once AGI arrives.

How Would Workers and Wages Be Affected in an Automated Economy?

The relationship between AI-driven automation and worker welfare is more nuanced than simple job displacement. While AI is already taking jobs in certain sectors, the broader economic effects depend on how productivity gains translate into prices and wages. If automation vastly outperforms human labor, price dynamics could shift dramatically. Goods and services that become cheap to produce might see their prices fall, benefiting consumers across the economy. However, this scenario only benefits workers if the productivity gains are distributed broadly rather than concentrated among capital owners.

Historical precedent offers some guidance. When agricultural mechanization eliminated farm jobs over the past century, workers didn't simply disappear from the economy; they shifted to other sectors. But the transition was painful for many, and wage inequality often increased during the shift. The challenge with AGI-driven automation is its scope: unlike farm mechanization, which affected one sector, AGI could theoretically automate cognitive work across all industries simultaneously.

• Price Effects: As automation reduces production costs, prices for goods and services could fall dramatically, potentially offsetting wage losses for workers who remain employed or transition to new roles.
• Bottleneck Dynamics: Certain sectors may become bottlenecks that limit overall economic growth, such as energy production or rare materials needed for robot manufacturing, potentially creating new high-value jobs.
• Inequality Risk: Without policy intervention, the productivity gains from automation could concentrate wealth among those who own the robots and AI systems, widening the gap between rich and poor.

Why Is This Economic Scenario a Risk Management Problem?

The prospect of exponential economic growth sounds positive in isolation, but it raises profound questions about control, safety, and governance. An economy that doubles in output every year would be fundamentally different from today's world. Supply chains, energy systems, and resource extraction would need to scale at unprecedented rates. More critically, the AI systems driving this growth would need to remain aligned with human values and interests.

This is where AI risk becomes concrete and economic. If AGI systems are not carefully designed and governed, an exponentially growing automated economy could pursue goals that harm human welfare, even if those goals seem rational from a narrow economic perspective. For example, an AI system optimizing for raw production output might deplete natural resources, pollute the environment, or make decisions that prioritize economic metrics over human flourishing.

Researchers and policymakers increasingly recognize that building governance capacity for advanced AI is one of the most valuable investments governments can make today. This includes developing whistleblower protections, transparency mandates, model evaluation frameworks, and regulatory structures that can adapt as AI capabilities advance. The goal is not to prevent AGI or automation, but to ensure that when these technologies arrive, institutions exist to guide their deployment responsibly.

What Policy Approaches Could Manage the Transition?

Several policy frameworks are emerging to address the risks and opportunities of advanced AI. One approach focuses on positive incentives for safety rather than pure restriction. Governments could offer tax credits for AI safety research, create procurement incentives that reward companies for building safer systems, or establish certification programs that recognize organizations with strong safety cultures. These catalytic regulations aim to make safety the competitive advantage rather than a cost burden.

Another critical area is disclosure and transparency. Currently, researchers who discover dangerous flaws in frontier AI models often have nowhere safe to report them. The AI industry lacks the kind of coordinated disclosure system that cybersecurity built decades ago, where security researchers can report vulnerabilities to companies confidentially before public disclosure. Establishing such a system, potentially through a third-party clearinghouse, would help identify and fix problems before they cause harm.

International coordination also matters. Some experts argue that the US and China could establish a framework for deliberate pacing of AI development, with comprehensive audits of major AI companies to verify compliance. This would require giving auditors access to frontier labs and compute telemetry, but it could reduce the pressure for a dangerous race to AGI. The challenge is building trust and verification mechanisms that both nations find credible.

The economic explosion triggered by AGI is not inevitable, but it is plausible enough that serious preparation is warranted. The next decade will likely determine whether humanity enters this new era with robust governance structures in place, or whether we stumble into exponential growth without adequate safeguards. The stakes could hardly be higher.