from Hacker News

Researchers experimented on Reddit users with AI-generated comments

by devrandoom on 4/29/25, 8:19 AM with 10 comments

• by IAmGraydon on 5/4/25, 6:57 PM

  Reddit is nothing more than a social manipulation platform, and their feigned outrage at this is hilarious. They are absolutely complicit in allowing this very thing by allowing signups without any identity verification. They don't stop it because they know that a large percentage of their traffic is bots, and doing so would negatively impact the value of the company. So once again, the right thing is set aside for money.

• by remram on 4/29/25, 3:49 PM

  I'm of two minds about this.

  On the one hand, this is egregious and should not happen. People should not be stalked and lied to, and communities should not be manipulated at this scale.

  On the other hand, it certainly already happens as it is incredibly easy for any group to use LLMs and do this. So the process and effectiveness should be studied and reported to the public... There might be a better way to do it though?

• by devrandoom on 4/29/25, 8:22 AM

  Reddit is furious, understandably.

  I wonder if other groups are using AI to post on Reddit too?

• by remram on 4/29/25, 5:02 PM

  Discussed here: https://news.ycombinator.com/item?id=43806940 (238 points, 168 comments)