Scientists Asked AI to Impersonate 112 Public Figures. What Happened Next Is a ‘Dire’ Warning
404 Media
A PLOS One study reveals that people often perceive AI-generated impersonations of public figures as more authentic and coherent than the actual individuals.Summary
Researchers utilizing GPT-4 Turbo found that AI-generated responses from 112 public figures were consistently rated as more authentic, coherent, and relevant than real debate responses from the BBC's "Question Time." Led by Steffen Herbold, the study suggests that AI's ability to convincingly mimic human speech poses a significant risk for political misinformation. Participants were largely unable to distinguish between human and machine, highlighting an urgent need for public awareness and potential regulatory frameworks to combat AI-driven deception.
(Source：404 Media)