The Future and Its Enemies

In the AI Story

The book emerged at the height of 1990s techno-optimism but refused easy triumphalism. Postrel sided with dynamism while documenting legitimate stasist concerns: disruption creates real casualties, experimentation produces genuine failures, decentralized systems generate outcomes no one designed and some people hate. Her dynamism was empirical rather than ideological—grounded in observation that centralized attempts to direct complex change consistently fail, not in moral conviction that markets are sacred. She distinguished her position from both libertarian dogmatism and progressive planning—a third way that treated openness as pragmatic rather than principled.

The framework's application to AI has been explicit and extensive. Helen Toner's May 2025 essay used Postrelian language to critique AI safety community assumptions—that fewer leading AI projects would be safer, that development should be concentrated and governmentally supervised, that nonproliferation is the path to security. Toner identified these as stasist positions producing stasist risks: concentration creates single points of failure, reduces competitive pressure for safety, and eliminates the decentralized experimentation that reveals problems central plans miss. The value of open models—diverse use cases, distributed testing, broad research access—is obvious through a dynamist lens.

Postrel's own 2024–2025 commentary has been cautiously dynamist. She highlighted critics of Biden's October 2023 AI executive order who called it 'premature political solution to unknown technical problems and clear case of regulatory capture.' The stasist impulse—regulate early, control tightly, prevent proliferation—has dominated government responses globally. The dynamist alternative—strengthen institutions, support transitions, invest in human capacity—remains underrepresented in policy discourse. Postrel's framework provides vocabulary for the position: not laissez-faire indifference but active investment in the conditions that make decentralized adaptation successful.

Origin

The book's genesis was Postrel's frustration with political discourse that classified every position as left or right when the actual fault line ran perpendicular. She observed environmentalists fighting environmental cleanup technologies, safety advocates blocking safety innovations, progressives defending incumbent industries against creative destruction—positions that made no sense along traditional ideological axes but were perfectly coherent along the dynamist-stasist axis. The framework emerged from taking political actors at their word about what they feared (uncontrolled outcomes) rather than reducing their positions to economic interest.

The intellectual lineage combined Hayek's dispersed knowledge, Jane Jacobs's organic urbanism, and evolutionary theory's preference for variation over design. Postrel synthesized these into a political orientation: trust emergence over planning, prize diversity over uniformity, prefer reversible experiments to irreversible commitments. The book was unusual in treating technological change as the lens through which political philosophy should be refracted—reversing the standard priority that treats politics as foundational and technology as applied context.

Key Ideas

Dynamism versus stasis as the fundamental axis. The most important political division is not left versus right but open-ended experimentation versus centralized control—a distinction that cross-cuts traditional ideology.

Knowledge is dispersed and tacit. The information required to direct complex systems is distributed across millions of actors and embedded in local context—making centralized planning structurally inadequate regardless of planners' intentions.

Experimentation requires tolerance for failure. The process that produces valuable innovations necessarily generates many failures—systems optimized to prevent failure eliminate the experimentation that produces progress.

Emergent order can be legitimate. Outcomes no one designed can be preferable to outcomes anyone designed—not always, not automatically, but frequently enough that the default should be openness rather than control.