Fetching from the wire…
01
02
03
04
05
06
07
08
Research2026-08-20 · source-backed
The pattern is: ask an LLM for numeric judgments like willingness-to-pay, fit a utility function, choose actions from it. That requires a single utility function to reproduce the judgments. This paper tested whether stated WTP differences between two items match the stated indifference payment. They don't. (arXiv 2608.17644) If you're building personalization on LLM-scored comparisons, the arithmetic underneath doesn't close.
Each link below shares sources, entities, or timing with this story.
LLM uses OpenAI / Shared entity: LLM / Earlier coverage / Tension
Linked by a graph relationship (LLM uses OpenAI); both cover LLM; earlier LLM coverage from 2026-07-27.
Simon Willison released LLM / Shared entity: LLM / Earlier coverage / Tension
Linked by a graph relationship (Simon Willison released LLM); both cover LLM; earlier LLM coverage from 2026-08-16.
Linked by a graph relationship (Simon Willison released LLM); both cover LLM; earlier LLM coverage from 2026-06-19.
Linked by a graph relationship (Simon Willison released LLM); both cover LLM; earlier LLM coverage from 2026-06-18.
LLM uses OpenAI / Shared entity: LLM / Earlier coverage
Linked by a graph relationship (LLM uses OpenAI); both cover LLM; earlier LLM coverage from 2026-07-31.
Linked by a graph relationship (LLM uses OpenAI); both cover LLM; earlier LLM coverage from 2026-06-19.
Simon Willison released LLM / Shared entity: LLM / Earlier coverage
Linked by a graph relationship (Simon Willison released LLM); both cover LLM; earlier LLM coverage from 2026-08-17.
Linked by a graph relationship (Simon Willison released LLM); both cover LLM; earlier LLM coverage from 2026-08-12.