Why This Matters

If AI models begin training on data generated by previous AI models, the speed of technological advancement could move from linear to exponential. This shift could decouple economic growth from traditional human labor constraints, fundamentally altering how capital is allocated in tech portfolios.

Frontier AI models are increasingly being utilized to develop their successor models, creating a potential feedback loop that could trigger an intelligence explosion. This shift represents a transition from human-led research to automated, self-improving systems.

Self-Improving Models Threaten to Break Traditional Growth Models

The ability of frontier AI models to develop their next iteration introduces a framework for understanding explosive growth through new feedback loops (VoxEU, CEPR). This mechanism allows for a massive boost to technological advancement by removing the human bottleneck in the research and development cycle. If these loops become self-sustaining, the speed of innovation may outpace current regulatory and economic frameworks.

The potential for this explosive growth depends on three critical variables: the labor share (the portion of income received by workers as wages), the returns to research effort, and the share of capital allocated to these systems (VoxEU, CEPR). If the returns to research effort are high enough, the loop accelerates. However, if the costs of compute (the processing power required to train models) scale faster than the intelligence gains, the loop may stall.

Tech leaders claim AI will lead to less work and more free time for the workforce (BBC Business). However, current data suggests a massive disconnect between corporate optimism and worker reality. Many staff members report working up to 90 hours a week (BBC Business) as they attempt to keep pace with the rapid deployment of these new tools.

Meta and Boeing Pivot to Open and Autonomous Strategies

Large technology firms are restructuring their core business models to capture the value of these accelerating feedback loops. Meta is betting heavily on Open AI—the use of open-source or publicly available model architectures—to catch up in the global race (NYT Business). This strategy aims to foster a wider ecosystem of developers to accelerate model refinement.

In the physical world, the shift toward autonomous systems is already manifesting through strategic divestment and equity swaps. Boeing is selling three of its autonomous flight subsidiaries to Archer Aviation (NYT Business). In exchange for these assets, Boeing will receive a 16.5% stake in Archer, which is currently developing piloted electric aircraft (NYT Business).

Meta vs. Boeing Strategic Shifts

Meta is prioritizing software-based ecosystem dominance through open models (NYT Business). Boeing is pursuing hardware-integrated autonomy through equity stakes in specialized aviation firms (NYT Business).

The Scramble for Critical Minerals Risks a New Resource Curse

The physical infrastructure required to power these intelligence loops relies on a highly concentrated supply chain. The demand for lithium, cobalt, nickel, copper, and rare earth elements is reshaping global trade and great-power politics (VoxEU, CEPR). This demand is reminiscent of the earlier scramble for natural resources, which often resulted in economic instability for extracting nations.

The current landscape is characterized by a supply chain that is bifurcated between scattered extraction sites and Chinese-concentrated refining (VoxEU, CEPR). This concentration creates a strategic vulnerability for nations attempting to build domestic AI hardware. The 'esource curse'—a phenomenon where countries with abundant natural resources experience stagnant economic growth—remains a significant risk in this new technological era (VoxEU, CEPR).

Economic Divergence and the End of Unfettered Globalization

The transition toward AI-driven growth is occurring against a backdrop of extreme global instability. The era of unfettered globalization and widely shared macroeconomic objectives is effectively over (Project Syndicate). Corporate executives and investors must now adapt to a perpetual, directionless transition (Project Syndicate).

This period of volatility is marked by shifting trade dynamics and the rise of economic statecraft. While some nations criticize trade surpluses, these may actually represent a transfer of purchasing power that can 'enrich' neighbors in a high-capacity economy (Project Syndicate). As AI accelerates productivity in some regions, the gap between high-tech hubs and resource-extracting nations may widen significantly.

If AI-driven intelligence loops decouple growth from human labor, how will tax structures evolve to maintain social stability?

Key Terms
  • Frontier AI Models — The most advanced, high-performance artificial intelligence systems currently in existence.
  • Labor Share — The total portion of a nation's or company's income that goes to workers as wages rather than to owners as profit.
  • Compute — The amount of computational power required to perform complex tasks, such as training or running an AI model.
  • Resource Curse — An economic phenomenon where countries with a high abundance of non-renewable natural resources tend to have less economic growth and less democracy.