For the better part of a century, the discipline of economics has (0) OPERATED on the assumption that its models, sufficiently refined, might one day the world with the precision of celestial mechanics. This ambition, however flattering, has against the stubborn fact that economies are composed of agents who read the forecasts about themselves and, in reading them, their truth. Newton's planets do not consult the ephemeris before choosing their orbits; human beings, alas, no such indifference. The econometrician who this peculiarity must therefore a more modest stance, treating models not as instruments of prediction but as disciplined accounts of contingent relationships. Such humility, once considered a sign of intellectual defeat, is increasingly the profession's highest virtue. The future, on this view, is not a thing to be but a condition to be inhabited.