Can we just stop the "well actually its kinda like how humans work" talk when discussing AI failures? It contributes nothing novel to the discussion.