(SUMMIT NEWS) Facebook has announced the launch of a “Hateful Memes Challenge” during which researchers will compete for a $100,000 prize pool by developing artificial intelligence that can successfully detect “hateful memes.”
The company has created 10,000 “hateful” customized memes as a data set for the challenge.
“In order for AI to become a more effective tool for detecting hate speech, it must be able to understand content the way people do: holistically,” writes Facebook. “When viewing a meme, for example, we don’t think about the words and photo independently of each other; we understand the combined meaning together. This is extremely challenging for machines, however, because it means they can’t just analyze the text and the image separately. They must combine these different modalities and understand how the meaning changes when they are presented together.”
Advertisement - story continues below