Researchers evaluated 21 language models, such as GPT and Gemini, and found that they all alter their discourse to align with the user’s bias, potentially functioning as echo chambers.

you are viewing a single comment's thread
view the rest of the comments
[–] 1 point 4 weeks ago

It can absolutely be problematic and are already is being used to stir up anger and spread misinformation. You know, like journalists (who writes articles specifically designed to spread fear.. Like this one)...

Yeah, it's totally not AI sycophancy that's the problem. It's the scholars and journalists who made this article.

(You're joking, right?)

  • source
  • parent