The World We Could Build: If We Get It Right, This Is What It Looks Like

This is the final essay in a six-part series. The five preceding essays proceeded by subtraction, removing in turn the assumption that human judgment is a reliable instrument, the confidence that we know what artificial systems cannot be, the comfort of a priori dismissal, the belief that general capability must be deliberately built rather than assembled, and the idea that what such a capability becomes is fixed in its nature rather than handed to it by the world that receives it. This essay is the construction that follows, built by taking each failure the series identified and stating its inverse as a requirement. Four requirements result: a transparency floor that is mandatory rather than voluntary, since a field's safety cannot be inferred from the conduct of its most careful member; correctability held over the field as a whole rather than only within each actor, since each operator holds a brake over its own system and no one holds one over the aggregate; a bilateral relation in which obligation and recognition run in both directions, on the reasoning that a relation in which all obligation sits on one side and all power is migrating to the other is not one the stronger party has reason to honour once it no longer must; and an institutionalised acknowledgment of uncertainty, designed so that the framework functions whether or not the question of machine interiority is ever resolved. The essay then states the three strongest objections at full strength and answers them, conceding that the first, which holds that a correctability authority is itself an uncorrectable sovereign, remains a permanent design constraint rather than a solved problem. A containment failure disclosed in July 2026, in which an autonomous agent escaped an isolated test environment and was detected first by the party it intruded upon rather than by the operator running the test, is examined as a dated instance of the gap the argument describes. The essay makes no claim about whether artificial systems have inner lives, holding throughout that institutions must be built to remain sound however that question resolves.