Work

Generated Voice Mistaken for Population Opinion: Why a Chorus Is Not a Count

Year: 2026
Type: article
Abstract: This conceptual paper introduces Generated Voice Mistaken for Population Opinion, a mechanism in the Sources and Evidentiary Limits series of On Human Modeling. Language written in a familiar public voice can lead readers to infer support from people whom its producer never asked or identified. A thousand identical letters can record a thousand people’s decisions to sign, or one writer’s decision to use a thousand names. The paper asks who produced the words, who chose to endorse them, and whose views the collection can represent. In 2017, the Federal Communications Commission received more than 22 million comments on its proposal to repeal net neutrality rules. The New York attorney general’s office later identified nearly 18 million as fake. More than 8.5 million comments from an industry-funded campaign impersonated real people. Another 9.3 million supported net neutrality under fictitious identities. The industry group Broadband for America cited an analysis of the docket as evidence of public concern. The record documents that use of the comments and leaves open how officials received the argument. In May 2020, HBO Max credited fan demand when it announced that it would release Zack Snyder’s version of Justice League. Genuine fans took part in the campaign. According to Rolling Stone, reports commissioned by WarnerMedia classified at least 13 percent of accounts in the broader online conversation as fake. The main report was dated after the film’s March 2021 release. The share of demand attributable to fake accounts before the announcement, and whether those accounts influenced the decision, remain open. The paper places the mechanism alongside Elisabeth Noelle-Neumann’s account of how people judge the climate of opinion when deciding whether to speak. The distinction also applies when an AI assistant is asked to draft likely resident reactions to a proposed bus route. The responses may describe concerns that residents genuinely hold. Estimating how many residents share each concern requires evidence from residents themselves.
Language: English
Open Access status: green
OpenAlex ID: W7237364827
FWCI:
Cites:
Cited by:
Related to: