
PDF is a typesetting format. Behind the scenes, every letter is placed on the page at a specific x,y coordinate. parsing text out of a pdf is essentially a series of guesses. sometimes the algorithm guesses wrong. PDF was invented for perfect layouts for printing.
Word docs are a text-based format. It’s very easy to correctly pull text from a Word doc.
Why risk an ai parsing a pdf incorrectly? There is no upside.
He just seems like a slimy shithead that was bullied constantly for being a prick to everyone. Now he wants revenge on everyone, becuase its their fault for not accepting his dickish behavior.