Comment by tibbar

I think Pangram is de facto measuring the default Claude voice. It doesn't fire for me on technical GPT writing.