At the same time, AI models that are trained primarily with text to detect hate speech, struggle to identify hateful ... model to detect hateful memes is a multimodal problem.

In social media, everyday speech, and, of course, Internet memes, we tend to see the phrase “first world problems” tossed around a lot. It’s a pretty humorous term referring to a problem so ...