Man trains a bot on 4-Chan.
He leaves it running and it produces a spew of bile and hatred. People are dismayed!
To me it is no surprise at all that a bot trained on an offensive messaging board reproduced offensive content. The boards are obviously pretty awful, but that is a problem as old as the internet. That it can be reflected back automatically by an ML trained bot is unfortunate but expected. Is this a potential problem for society? Could bad actors unleash hordes of nasty bots? Possibly, but I think humans are doing a pretty good job of this themselves.