I had the same thought, but I think this except from the post assuaged some of those concerns:
"We want to make clear here that even though the current version of the robot appears to be performing amazingly well, our human flaggers are just as important as ever. The only reason the robot is able to do as well as it’s doing is because of the 100,000+ flags that humans have raised and moderators have handled. The robot is a way to find things that look like things humans have flagged before, but we don’t think it can identify truly novel modes of unfriendliness. Only the humans in our system can do that. So if you see unfriendliness in the comments, please flag it. You’re helping to make Stack Overflow better."
Unfriendly comments are the least of their problems. The fundamental problem is their broken reputation system.
Questions are outright closed by people with a high enough karma even if it is outside of their knowledge domain. People who leave helpful feedback after a downvote will receive a retaliatory downvote in turn. People with massive karma will receive upvotes even if their answer is not a good one. To boost karma, it's better to be first than good, in general. New questions are closed and redirected to old questions with outdated answers. Non-native English writers get downvoted for grammatical errors, even if the question is fine. One downvote will result in more, as the mob piles on. Bringing up any of these issues in their "meta" results in downvotes.
So, Stack Overflow will focus on its moderation system and "unfriendly comments" while its entire system encourages a kind of arbitrariness that is itself unfriendly.
Exactly, the biggest problem is the culture of existing users, which was built up by stack overflow encouraging a competition for internet points which leads to all kinds of antisocial behaviour and attempts to game the system and discourage participation by rivals.
"We want to make clear here that even though the current version of the robot appears to be performing amazingly well, our human flaggers are just as important as ever. The only reason the robot is able to do as well as it’s doing is because of the 100,000+ flags that humans have raised and moderators have handled. The robot is a way to find things that look like things humans have flagged before, but we don’t think it can identify truly novel modes of unfriendliness. Only the humans in our system can do that. So if you see unfriendliness in the comments, please flag it. You’re helping to make Stack Overflow better."