Huh? Multiple models already have been sourced from quality data that's not scraped from the open web.

→ View original post on X — @alexjc