Model ownership and data sovereignty—this is the real way forward. Imagine if the community could truly own the intelligent systems, while individuals fine-tune models with default privacy, then we could have truly personalized AI without sacrificing identity information and privacy. Sounds good, right? But the challenge lies in how to design incentive mechanisms and governance systems. How can we motivate all participants to contribute? Who makes the decisions? These are the real problems that need to be solved.