>Arguably, AI hypers should lead with data not with anecdotal evidence

This reminds me of people who get sad when they realize they haven’t discovered anything amazing.

I am pedantic and “people that” → “people who” (for people, who is preferred).