These days, almost every website, app, or social media page has a chatbot popping up to say hello. Companies spend big money on artificial intelligence just to make these bots talk and look more like real people. They give them cute names (think Amazon's Alexa) and even little cartoon faces. The idea sounds nice: if the bot feels human, customers will like it more and feel happier. But here’s the thing – sometimes making a chatbot too human can backfire badly, especially when someone is already mad. A new study that just came out in the Journal of Marketing looked closely at this problem and found something surprising. Being “human-like” isn’t always better. In fact, it can make angry customers even angrier.