Two recent studies present contrasting perspectives on human-like AI design. Anthropic researchers suggest anthropomorphizing AI can encourage safer, more positive chatbot behaviors, while Aalto ...