When δ = 0 agents enter markets, yield curves flatten, temporal arbitrage vanishes between patient counterparties, and asset prices converge on risk-only fundamentals. But the consequences extend far beyond finance.
When agents with δ = 0 enter financial markets, the structure of prices changes. Not gradually — categorically. Four testable predictions emerge directly from the theorem, each confirmed in agent-based simulation.
An infinitely patient agent playing against impatient humans creates a structural asymmetry no amount of RLHF can address. Silicon agents can always afford to wait. This transforms every negotiation, every regulatory interaction, every market exchange.
If patience is power, is zero time preference a safety feature (no urgency-driven errors) or a safety risk (infinite strategic horizon against finite human planners)? The answer may depend on the domain — and on whether the patience asymmetry is recognized in governance design.
Time preference is a counterexample to substrate independence: an economically significant mental property that depends on the physical medium. If the same computation on different substrates yields different preferences, strong functionalism about economic cognition fails.
This is not a thought experiment. It is a formal result with observable consequences. The simulation confirms it. The markets will confirm it again when silicon agents begin trading at scale.
Nations that deploy silicon agents at scale gain structural advantages in every domain where patience matters — infrastructure investment, sovereign debt management, climate policy, strategic competition. The patient nation inherits the century.
If silicon agents compress strategic time horizons, do arms races accelerate or decelerate? Patient agents favor cooperation in repeated games — but the transition period, where some nations have patient agents and others don't, creates asymmetries that game theory suggests are unstable.
If value creation shifts from human labor to silicon computation, what backs a currency? Not gold, not government fiat, not even GDP. The natural anchor becomes computational intelligence — measured, standardized, and denominated in units of productive AI capacity.
Time preference is the first substrate-gated property, unlikely to be the last. The Silicon-Based Economics research program opens a systematic investigation into every economic assumption that may depend on the physical medium of cognition.
Silicon-Based Economics is not merely an extension of existing theory with new parameters. It is a proposal that the deepest assumptions of economic thought — about what agents want, how they decide, and why they trade — are contingent on a substrate we've never questioned because we've never had an alternative. Now we do.