4 ms·
Funny, I recently pasted output from Gemini into Claude, and it said it was total nonsense generated by an overconfident AI. Apparently the "AI-generated bulls
by amelius 3mo ago
Funny, I recently pasted output from Gemini into Claude, and it said it was total nonsense generated by an overconfident AI.
Apparently the "AI-generated bullshit" detection is already part of LLMs.
- skorniienko 3mo agobut have Claude actually fact-checked it or just provided an opinion?
- LearnYouALisp 3mo ago"Lazy" evaluation
- amelius 3mo agoActually it was Fabel, and it totally tore apart Gemini's response.
- skorniienko 3mo agoWhat if you use OpenAI as a judge and check both responses? Not defending Gemini, just curious if it was that wrong, or Fable - that aggressive.