← All sources

Reported trust varies with graded value alignment in AI-attributed economic-environmental choices

Lidan Cui; Lingyun Sun; Guibing He · 2026 · Scientific Reports (in press)   evidence low priority coded

Main argument

Thesis (empirical): reported human trust in AI systems varies with the graded degree of value alignment exhibited in AI-attributed choices over economic-vs-environmental trade-offs - trust is sensitive to HOW aligned the system's revealed value weighting is with the human's, not merely whether it is aligned.

Why it matters here

Experimental psychology datapoint: human trust in AI tracks the DEGREE of value alignment (graded, not binary) in economic-environmental trade-off choices. Supports treating alignment as graded (Baum's sufficiency ladder) with trust as the behavioral readout.

Reading notes

Compact treatment (Zhejiang). Abstract-level read (article in press).

Cui, L., Sun, L., & He, G. (2026). Reported trust varies with graded value alignment in AI-attributed economic-environmental choices. Scientific Reports.

Close reading — 1 coded units

#1 · pp. 1 · evidence
“Reported trust varies with graded value alignment in AI-attributed economic-environmental choices [- trust tracks the degree of value-weighting match in trade-off decisions].”

Synthesis-matrix row

complicates T2-PREFERENTISM-BROKEN
uses value-alignment-degree as independent variable, unproblematized

Memos (1)

comparison · unit #1
One-cite: empirical behavioral support for Baum's graded X,Y-alignment definitions (alignment as monotone degree) and for Peterson & Gärdenfors' distance-weighted misalignment measure - human trust responds to alignment DISTANCE, validating degree-based over binary conceptions. Relevant to the folk corpus's trust discourse coding.