A new study reveals how large language models (LLMs) respond to different motivational states. In their evaluation of three LLM-based generative conversational agents (GAs)--ChatGPT, Google Bard, and ...