In tests involving the Prisoner's Dilemma, researchers found that Google’s Gemini is “strategically ruthless,” while OpenAI is collaborative to a “catastrophic” degree.
  • Johannes Jacobs@lemmy.jhjacobs.nl
    4·
    2 days ago

    With coding, i have the opposite idea. ChatGPT would often say “oh noes! Cant do that because it might be harmful!” While Gemini happily chirps out the code, and simply warns me of any dangerous actions

  • FundMECFS@quokk.auEnglish
    8·
    4 days ago

    This actually matches my experience. Chatgpt will talk to me about dismantling the state if I tell it to be anarchist while Gemini will lecture me about a “extremist ideology”