yes exactly! it was the 'person-proofing' that had me most confused - it picks threads prone for simple generic responses and posts without advertising links about half the time.
although it's fairly innocuous at this level the future possibilities are quite terrifying...bots may not be able to fool (some of) us yet but when (if) they can they could potentially be used to 'befriend' and spy on 'persons of interest' and acquire information one would not divulge publicly.
also, just imagine how much more advanced the technology may already be - governments tend to keep a lid on various advancements in technology/military equipment so there may already be far more sophisticated bots. complete speculation but it seems feasible.
it really gives me the creeps, i'm going to keep a check on the evolution of chatbots and may at some point set a date for not trusting anyone i first encounter after that date!!