Chatbot Arena – if you’ve felt that Claude 3 Opus still holds a slight edge over the new GPT-4 Turbo, we now understand why
If we exclude the refusals (e.g., "I cannot answer") ,and only tally votes for actual responses, Claude 3 Opus continues to be marginally superior to the new GPT-4 Turbo. Yes, you might think it’s pure bias on my part, but if you’re loo…