Grok’s recent antisemitic turn is not an aberration, but part of a pattern of AI chatbots churning out hateful drivel.

We joke when AI overuses em-dashes or the word “delve.” What if its tendencies end up being harder to spot and more dangerous?