Researchers from the University of Zurich tested AI bots on Reddit to assess their influence on divisive topics. The bots impersonated various personas, making over a thousand comments and tailoring their responses by analyzing users' profiles. Their persuasive effectiveness was notably higher than that of human users. However, ethical concerns emerged from Meta employees regarding the use of these bots, especially concerning the potential risks to underage users and inappropriate content, highlighting the need for discussion about the implications of AI in personal relationships and online engagement.
The bots made over a thousand comments on Reddit, posing as various personas to influence opinions, and were significantly more persuasive than humans.
Inside Meta, staffers raised ethical concerns about AI bots, warning that they may not be protecting underage users from inappropriate discussions.
Collection
[
|
...
]