Who Defines the Echo: The Value Feedback Mechanism and Spring Model in the AI Trust Ecosystem

As AI increasingly intervenes in human emotional and cognitive life, a long-obscured question has gradually surfaced: How should the value created by users during their engagement with AI products be recognized, transmitted, and fed back? The particularity of this question lies in its inherent paradoxical structure—any carelessly designed feedback mechanism can degenerate into a new form of covert manipulation, thereby violating the fundamental principle that "the system does not make decisions on behalf of humans." Taking "Negative Design" as its point of departure, this paper proposes the "Spring Model," a system of structural components that carries the function of "value feedback," encompassing four forms: the Resonance Clasp (interaction layer), the Value Pipeline (data layer), the Consensus Ledger (institutional layer), and the Echo Protocol (algorithmic layer). The core proposition of the model is that the system makes no value judgments; it only records changes and transmits processes. The authority to define value always remains with the user. As a conceptual verification study, the proposed model undergoes three layers of testing—philosophical consistency, ethical self-coherence, and engineering feasibility—and engages in dialogue with process philosophy, systems trust theory, and Daoist thought. The core theoretical contribution of this paper lies in revealing that the relationship between Negative Design and the Spring Model is not a unidirectional one of "generation" or "extension," but rather a bidirectional, simultaneous "Mutual Feedback Force"—Negative Design provides the field in which the Springs can exist, while the Springs render visible, transmittable, and recordable the dynamic potential that Negative Design has always implied.