Open Menu
AllLocalCommunitiesAbout
lotide
AllLocalCommunitiesAbout
Login

OpenAI is trying to clamp down on ‘bias’ in ChatGPT

⁨65⁩ ⁨likes⁩

Submitted ⁨⁨5⁩ ⁨days⁩ ago⁩ by ⁨inclementimmigrant@lemmy.world⁩ to ⁨technology@lemmy.world⁩

https://www.theverge.com/news/798388/openai-chatgpt-political-bias-eval

source

Comments

Sort:hotnewtop
  • MaggiWuerze@feddit.org ⁨5⁩ ⁨days⁩ ago

    lol, so we’re gonna get another Hitlerbot in their attempt to move the bot away from the reality it learned from its training data?

    source
  • MotoAsh@piefed.social ⁨5⁩ ⁨days⁩ ago

    Too bad for them, but reality has a left bias when the right denies the basics of reality… All such “bias removing” training will do is make a worse product.

    source
  • HuskerNation@lemmynsfw.com ⁨5⁩ ⁨days⁩ ago

    Can we stop pretending that truth and facts are biased. If you don’t like it fucking cope and seed

    source
  • hedge_lord@lemmy.world ⁨4⁩ ⁨days⁩ ago

    Clamping down on the affirmation bias right? Right??

    source
    • recked_wralph@lemmy.world ⁨4⁩ ⁨days⁩ ago

      That’s a great point and well articulated! Let’s see if we can expand on this together

      source
    • Squizzy@lemmy.world ⁨4⁩ ⁨days⁩ ago

      Great idea - impressive comprehension from associated reports

      source
  • SubtitledMustard@lemmy.world ⁨4⁩ ⁨days⁩ ago

    Let’s call it what it is. Political Correctness. Time to start using the coded language mantras conservatives used to use back at them.

    source
  • phutatorius@lemmy.zip ⁨4⁩ ⁨days⁩ ago

    Did anyone ask Altman if Trump had a diaper change before Sam licked it? My guess is not.

    source