Clockless Consensus - Truth-Selection via an Unknown-Future Schelling Point

Many of the questions that matter are empirical yet unanchored: there is a fact of thematter, but no oracle will rule on it and no one can name the day the evidence lands —where a disease outbreak began, what a dietary pattern does over decades. Predictionmarkets cannot reach them, because they need a settlement event to import. What remainsmeasures what a crowd believes now, and nothing in it pulls that belief toward evidence.Nor does adding a stake supply the pull: where belief is traded with no resolution toanchor payoff, momentum dominates and the profitable move is to front-run the crowdrather than to be right. We introduce clockless consensus, which supplies the missing pull without importing ananchor. A participant stakes real value on a True/False position; the position locksand settles neither on a date the holder picks nor at an oracle's ruling, but at a momentset by the activity of the question itself — more contested questions settle later,and no one can time their own settlement. Each position is thus its own syntheticresolution event, never a shared date. Reward rises with how lopsided the crowd wasagainst a position at entry, and with whether the crowd has come around by the time it ismeasured: bold at entry, correct at settlement. The conceptual core is an unknown-future Schelling point: a Keynesian beauty contestconverges only because every player is scored at one shared instant, and removing thatinstant leaves nothing to coordinate on except an answer that does not move — for anevidence-bearing question, the truth. We propose no theory of truth and do not claim thatperfect epistemic consensus is attainable; this is an incentive mechanism meant to improvethe long-run aggregation of dispersed evidence under uncertainty. We give the designrationale, a game-resistance analysis, and — centrally — a precise account of whysimulation can validate the mechanism's engineering properties but not its epistemicclaim.