Some interesting analysis in this piece. The short version is: categories that have been historically prone to falsity and scams (prior to AI) continue to be so, but at a more industrial scale.
The shorter version is: don’t get health advice from random people online.