Comment by b800h

I'm genuinely surprised that more people - including this mathematician in particular - don't untick the "improve the model for everyone" box. Unless the suggestion is that OpenAI ignore this preference?