Comment by NitpickLawyer

I have nothing against researching this, I think it's important. My main issue is with articles choosing to grab a "conclusion" and imply it extrapolates to larger models, without any support for that. They are going for the catchy title first, fine-print be damned.