Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Malicious AI swarms can threaten democracy (science.org)
6 points by imiric 6 days ago | hide | past | favorite | 4 comments
 help



This is already happening. Twitter / X is a great example. Look at replies to posts by Elon. Lots of accounts that use the same talking points, reply to each other, retweet each other, and are often retweeted by Elon. It’s manufactured dialogue that crowds out real replies by humans.

True, this has been a problem for decades, but this new technology makes the work required to spread disinformation and steer online conversations much more accessible, with far greater reach and impact. It's no longer necessary to run human troll farms, when anyone can deploy thousands of bots cheaper and faster. We haven't even begun to address the original problem, and we've now made it much worse. It's bewildering that very few people are taking this seriously.

I'm not sure this works forever. A big chunk of the real humans will eventually figure out that the real human conversation is somewhere else. At that point, X loses much of its influence. Sure, you can - with great effort - shout louder than all the bots with opposing views. So what? Nobody real is listening.

You're overestimating the capability of the average human to discern fact from fiction, and underestimating our susceptibility to disinformation. This is a problem even for educated skeptics. If all your information sources are flooded with disinformation, eventually you'll start paying attention to it.



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: