Can Llms Capture Human Preferences