Both of these terms are nothing new.

Snapped from the wild commit de9695dd • image visual-slop mystery-content 2 options extract text from image only and use that or fullslop analysis like image submit