A Reddit user argues that the popularity of “humanlike” Qwen models reflects prompting rather than specialized fine-tuning, claiming system prompts can make most models behave like conversational partners. They recommend defining a role and character background through a “You are...” prompt to shape the interaction.