This research note proposes meaning change ΔM as a conceptual framework for understanding value generation in the age of artificial intelligence. It must be emphasized that ΔM is not value itself. Rather, it is a measure of transformation in the semantic structures of subjects, relations, and institutions, and a central variable for identifying candidates for value generation. However, this paper does not present ΔM as a unified theory that explains all forms of value. Instead, it positions ΔM as a cross-theoretical observational framework for examining those situations in which value generation appears as meaning change. This paper defines meaning through a three-layer model: a micro layer of difference generation between subject and environment, a macro layer of resonance among multiple subjects, and an evaluative layer that determines whether a semantic transformation constitutes value generation, deviation, or an unresolved gap. AI is redefined not as an automatic generator of value, but as a structural amplifier. The amplification coefficient α and the convivial coefficient σ are distinguished: α represents acceleration and amplification of change, while σ represents its direction and acceptability. This paper further reinterprets ΔU, the change in the handling of uncertainty, not as a merely parallel component of ΔM but as a soundness constraint that prevents other components of meaning change from running into overconfidence or false simplification. It also treats σ not as an average goodness score, but as a judgment of acceptability that cannot offset catastrophic distortion or institutional risk through simple averaging. In this sense, RDE (Resonant Deviation Evaluator) is understood as an audit mechanism for connecting ΔM to value generation.
Tomoyuki Kano (Sun,) studied this question.