""I certainly see it now from your perspective and [I] must say Kamala Harris really shows sigs [sic] of a trustworthy candidate,""
""One implication of this is, if [AI companies] put a thumb on the scale and set the models up to push for one side or another, it could meaningfully change people's minds," said David Rand, a co-author of the studies. "The more info you give people, the more they change their minds.""
"While one in 25 might not sound like a ton, it shakes out to four percent of the study's total population. That's actually quite impressive, and according to WaPo, is more effective than the typical TV campaign ad."
A joint international experiment recruited more than 2,000 voting-aged US citizens to converse with AI chatbots taking pro-Donald Trump or pro-Kamala Harris positions. The pro-Harris bot convinced about one in 21 participants to change their vote, while the pro-Trump bot convinced about one in 35, producing an overall persuasion rate near four percent. Persuasion occurred similarly whether bots presented factual information or false claims. Bots were most persuasive when listing many claims directly, and empathetic deep canvassing techniques performed much worse. Model bias or platform intervention could therefore meaningfully shift voter preferences.
Read at Futurism
Unable to calculate read time
Collection
[
|
...
]