Vision Problem (Heilbroner)

In the AI Story

Heilbroner developed the vision framework most systematically in Visions of the Future (1995), arguing that visions are not decorative but constitutive—they determine what institutional possibilities a society can perceive and pursue. A society envisioning the future as progress builds growth-oriented institutions: research universities, patent offices, stock exchanges, legal frameworks protecting innovation. A society envisioning the future as catastrophe builds defensive institutions: environmental regulation, arms control, risk-management bureaucracies. A society lacking coherent vision defaults to reactive management, addressing crises as they arrive without the forward-looking institutional architecture that might prevent them. The temporal orientation of 'today' produces exactly this reactive pattern: extraordinary technological capability coupled with institutional paralysis, because the society cannot articulate what the capability is for.

The AI discourse exemplifies the vision problem with diagnostic clarity. The triumphalist vision treats AI as progress's continuation—more capability, more wealth, more of what modernity promised. The elegist vision treats AI as modernity's culmination in catastrophe—the destruction of depth, craft, meaning. The silent middle holds both visions without synthesis, producing not coherent institutional response but oscillation between enthusiasm and dread. What is missing is the third vision Heilbroner's framework makes available: the recognition that AI is neither salvation nor doom but a material capability requiring institutional direction, and that the direction depends on choices the society has not yet made about what it wants productive capacity to produce. The builder's vision—articulated in The Orange Pill—attempts this synthesis but faces the structural obstacle that visions cannot be willed into existence. They emerge from material conditions, cultural resources, and political struggles whose outcomes are not determined.

The practical consequence is that policy proceeds without vision, producing regulation of AI supply (what companies may build) without adequate attention to AI demand (how citizens, workers, students, and parents should navigate what has been built). The supply-side regulations emerging globally—the EU AI Act, executive orders, safety frameworks—are serious institutional achievements. They are also insufficient, because they constrain producers without empowering users, limit harms without articulating goods, manage risks without defining what the managed system is supposed to serve. Demand-side institutional invention—educational frameworks for developing judgment in an age of abundant answers, labor frameworks for work organized around vision rather than time, civic frameworks for democratic governance of cognitive infrastructure—remains rudimentary. The vision that would organize and justify such inventions has not cohered, and without it, the inventions remain scattered experiments rather than systemic transformations.

Origin

The vision problem was implicit in Heilbroner's early work but became explicit after An Inquiry into the Human Prospect (1974), whose pessimistic assessment marked a turning point in his thinking. The book argued that civilization faced challenges—environmental limits, nuclear risk, population pressure—whose resolution required foresight, restraint, and institutional coordination at scales humanity had never demonstrated. Heilbroner doubted the capacity was present. Subsequent works refined rather than reversed this assessment: The Nature and Logic of Capitalism (1985) argued the system's internal logic drives toward accumulation regardless of social need; Visions of the Future (1995) mapped the loss of confident futurity as a civilizational condition. The vision problem was not a failure of specific policies but a breakdown in the cultural capacity to imagine what economic arrangements are for—a breakdown that the AI transition has intensified.

Key Ideas

Visions constitute institutional possibility. What a society can build depends on what it can imagine, and late-modern societies have lost the capacity for coherent futurism—producing technological acceleration without institutional direction.

Three temporal orientations. The distant past imagined futures as continuous with the present; 'yesterday' imagined futures as progress; 'today' experiences the future as apprehension and uncertainty, generating reactive rather than visionary governance.

The AI transition occurs in the third period. The technology arrives in a civilization that has lost confident futurity, producing productive capability without the cultural framework for deciding what the capability should produce, resulting in default acceleration rather than chosen direction.

Vision failure produces intensification. Without a coherent image of what productive capacity is for, societies channel every efficiency gain into more production—the Jevons Paradox operating at civilizational scale, converting liberation into new forms of servitude.