Comment on AGI achieved 🤖

• LodeMike@lemmy.today ⁨3⁩ ⁨weeks⁩ ago

  Man AI is ass at this

  *laugh track*

  source
• rumba@lemmy.zip ⁨3⁩ ⁨weeks⁩ ago

  ohh god, I never through to ask reasoning models,

  DeepSeekR17b was gold too

  Image

  source

  • rumba@lemmy.zip ⁨3⁩ ⁨weeks⁩ ago

    then 14b, man sooo close…

    Image

    source

    • Mwa@thelemmy.club ⁨3⁩ ⁨weeks⁩ ago

      I wonder how QWEN 3.0 (14B AND 17B) performs

      source

      • rumba@lemmy.zip ⁨3⁩ ⁨weeks⁩ ago

        I don’t have any other models pulled down, if they’re open I’ll try it and respond back here

        source

        • -> View More Comments
    • merc@sh.itjust.works ⁨3⁩ ⁨weeks⁩ ago

      And people are trusting these things to do jobs / parts of jobs that humans used to do.

      source

      • Jakeroxs@sh.itjust.works ⁨3⁩ ⁨weeks⁩ ago

        Humans are pretty dumb sometimes lol

        source

        • -> View More Comments
  • ICastFist@programming.dev ⁨3⁩ ⁨weeks⁩ ago

    Oh god, the asses multiplied 🤣🤣🤣

    source
  • localhost443@discuss.tchncs.de ⁨3⁩ ⁨weeks⁩ ago

    It’s painful how Reddit that is…

    So,

    Now,

    Alright,

    source

    • rumba@lemmy.zip ⁨3⁩ ⁨weeks⁩ ago

      Probably where 90% of the training for this particular problem came from.

      source