The Long Council

Who was selected, and why

Can capitalism survive an AI-driven economy?

The panel · 16 June 2026 · 1 voices

The central tension

The live disagreement is not whether AI will create disruption — that is largely settled — but whether the disruption is a solvable distributional problem within capitalism's existing logic (requiring reform of ownership, taxation, and redistribution) or a structural crisis that exposes capitalism's internal contradictions (in labour value, price formation, and political legitimacy) that cannot be resolved without transformation of the system itself. **Pole A — Reformable capitalism:** AI represents a powerful productivity shock, but capitalism has absorbed previous shocks of this kind; the distributional consequences require policy intervention (taxation, basic income, retraining, antitrust) but not systemic transformation. **Pole B — Structural crisis:** AI is qualitatively different from prior technological shocks because it attacks the *mechanisms* capitalism depends on — labour as the source of value, price signals as information aggregators, distributed ownership as a legitimating condition — and the system cannot simply redistribute its way out of that.

Selected members

1. John Maynard Keynes

Aggregate DemandActive Fiscal PolicyManaging Uncertainty

Will argue: AI-driven productivity growth is capitalism's oldest promise finally delivered, but the distributional mechanism will not self-correct — aggregate demand will collapse unless governments actively redistribute the gains, likely through some form of guaranteed income and reduced working time; this is a demand-management problem, not a systemic-collapse problem, and capitalism's institutions are flexible enough to survive it if governments act counter-cyclically and with sufficient ambition. --- **2. Friedrich Hayek**

Keynes directly and explicitly addressed the question of technological unemployment and long-run distributional consequences of productivity growth, making him the council's most directly documented voice on this exact issue. · "Economic Possibilities for Our Grandchildren" (1930) is a direct documented engagement with the prospect of machine-displaced labour; his uncertainty framework (Treatise on Probability, General Theory Chapter 12) is directly applicable to the AI investment cycle and asset-price dynamics; his Bretton Woods work addresses concentration of surplus that AI ownership may reproduce at global scale.

Considered but not selected

*Amartya Sen** — directly relevant on the capability dimension (what AI does to what people are able to do and be) and on the democracy-as-epistemic-instrument argument. Not selected because the council already covers the distributional question (Keynes, Friedman), the structural-crisis question (Luxemburg), and the institutional-decay question (Hirschman); Sen's capability framework, while genuinely relevant, would partially duplicate the distributional-reform pole without adding a distinct line of argument. The more urgent gap is on Pole B, and Sen sits closer to Pole A (reformable, through capability investment). He would be the first addition if the session required deeper development-economics grounding.

*Elinor Ostrom** — initially considered because AI training data, algorithmic infrastructure, and digital commons are plausibly common-pool resources facing collective-action problems analogous to her studied cases. Not selected because the council's selection rules require that Ostrom's commons framework add genuine analytical traction to the *central tension* — which is about capitalism's systemic viability, not primarily about resource governance. The AI-as-commons framing, while interesting, is a secondary question; her framework would redirect rather than resolve the central debate. She would be the right selection for a session specifically on AI governance architecture or data commons regulation.

*John Rawls** — his difference principle (inequalities are only justified if they benefit the least advantaged) is a direct analytical tool for evaluating AI-driven inequality. Not selected because his framework is ideal theory operating at the level of principles, and the session's central tension is about the empirical and structural viability of capitalist institutions, not about the normative principles that should govern their design. Rawls provides a benchmark; he does not help answer whether the system can reach it. He would be valuable in a session that has already established the structural question and is now addressing the normative architecture of the successor order.