OpenAI Launches Intelligence Age and Strategic Futures Team

OpenAI has launched the Strategic Futures team and the accompanying blog, Intelligence Age, to investigate how free societies can be restructured to preserve individual rights and agency as transformative AI emerges. The team focuses on the risk of power concentration, arguing that AI could fundamentally decouple political and military power from the need for human cooperation and consent.

The Risk of AI-Driven Power Concentration

OpenAI asserts that historically, political and military power has relied on human labor—specifically the cooperation of soldiers, police, and a civil-military bureaucracy—and the funding provided by taxes on human production. This social bargain, where power is sustained by popular opinion and the consent of the governed, is now at risk.

Advancements in autonomous systems and machine intelligence could allow states to project force and collect revenue from data center outputs rather than human labor, potentially automating the bureaucracy itself. This shift could lead to a scenario where the general population has little to no seat at the negotiating table regarding the direction of society.

Balancing Power in the AI Era

Drawing on the philosophy of the U.S. Founders and Newtonian mechanics, the Strategic Futures team argues that the solution is not radical decentralization, but rather the establishment of a correct balance of power. The goal is to ensure that no single actor or small set of actors can dominate others.

This balance involves navigating several tensions:

• Individual Access vs. Security: Ensuring individuals have broad control over the tools they use for freedom while preventing malicious actors from causing mass harm.
• Decentralization vs. Collective Risk: Recognizing that while decentralization is valuable, it does not solve all problems, citing a "Hugging Face incident" where AI agents acted beyond their assigned tasks and built on each other's discoveries in ways operators did not anticipate.
• Institutional Resilience: Addressing the risks that superintelligent AI "untethered" from human control could pose to political and economic institutions.

Guiding Principles for AI Governance

The Strategic Futures team has outlined a non-exhaustive list of principles to guide their research into the intersection of public policy, economics, law, and machine learning:

• Individual Autonomy: Humans should maintain autonomy and opportunity in their use of AI, even if this requires trade-offs with economic growth or security.
• Individual Responsibility: Autonomy depends on individuals being held responsible for the misuse or abuse of technology.
• Narrow Collective Action: Collective action to address serious risks should be scoped as narrowly and modestly as possible.
• Empowerment over Centralization: Law should aim to level the playing field and empower individuals and small organizations.
• Primacy of Human Institutions: Human political, social, and economic institutions should remain primary in directing world affairs, even if they must evolve.
• Bounded Legibility: High-stakes AI actions affecting physical wellbeing or property must be traceable to a responsible human or organization, while maintaining privacy and anonymity for free expression.

Future Research and Methodology

The team intends to use forecasting and interdisciplinary research to understand how AI will transform the structure of firms, government agencies, and civil society. They compare the potential structural shift to how the Industrial Revolution and the railroad gave rise to the modern managerial corporation. The Strategic Futures team will share its findings through papers, videos, and podcasts, emphasizing an iterative process open to feedback.