Hacker News Reader

but have Claude actually fact-checked it or just provided an opinion?

Actually it was Fabel, and it totally tore apart Gemini's response.

What 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.

"Lazy" evaluation