There's a solution to this: stop linking to science journalism pieces. Instead, we should link directly to the paper.