中文版请点击这里阅读：https://sjr.sg/2026/01/02/16684/
Mathematics, logical reasoning, creativity, and perseverance—junior reporter programs serve as one of the best practical training models for the AI era; policymakers urged to strengthen scheduling, welfare, and accountability mechanisms to protect students and enterprises.
On New Year’s Day, the global wave of artificial intelligence continues to surge. China’s top AI research institution, DeepSeek, published an academic paper titled “Manifold-Constrained Hyperconnectivity: A Geometric Framework for Efficient and Interpretable Deep Learning” (abbreviated as mHC), drawing widespread attention in the international AI community. The paper introduces a novel computational framework based on modern differential geometry and topology principles, optimizing neural network connectivity to achieve more efficient knowledge transfer and sharing.
Although the paper is highly technical and difficult for general readers to fully grasp, its impact is already evident. DeepSeek has established joint laboratories with MIT and the University of Cambridge, focusing on the theoretical expansion of mHC; on GitHub, related open-source projects are growing exponentially. In industry, ByteDance explored this direction early on, and Google has also set up dedicated research teams to follow up.
How can junior reporters understand this technology?
One metaphor helps illustrate it: traditional AI models are like a single-lane highway, where information flows only in one direction; mHC is like widening the road into a multi-lane expressway equipped with intelligent traffic signals. Without requiring stronger hardware, it optimizes network connections mathematically, allowing information to flow more efficiently and intelligently.
There are differing views in academia regarding mHC’s future applications. Some scientists question its scalability for very large models, and manifold learning may unintentionally embed biases present in data. A report from MIT highlights a striking statistic: despite $30–40 billion invested by enterprises in generative AI, 95% of organizations see no returns. Ilya Sutskever, former Chief Scientist of OpenAI, believes the bottleneck in AI has shifted from computing power to innovative ideas.
Educators note that this trend poses new requirements for students, especially junior reporters: a solid foundation in mathematics remains essential, while language expression, logical thinking, creativity, and perseverance are equally crucial. Junior reporter programs serve as an ideal platform to train these comprehensive abilities. Through interviewing and reporting, students develop mathematical reasoning, logical judgment, communication skills, creativity, and the perseverance needed to tackle challenges—all while understanding the logic and reasoning behind technologies. For this reason, such programs are considered one of the best practical training models for comprehensive abilities in the AI era.
Teacher Dai Kai explains: “Local students may recognize the term ‘differential geometry,’ but truly understanding its meaning requires active exploration. Mastering foundational theories is not for technology’s sake, but to expand one’s thinking based on prior knowledge.” She emphasizes that AI is a tool, and human intelligence remains central, because no matter how advanced AI becomes, it still relies on humans to pose questions, issue instructions, and make decisions.
She further points out that the Ministry of Education should pay attention to scheduling and welfare for junior reporter programs and registered Chinese-language instructors, strengthening communication channels, rather than shifting responsibility to schools and leaving professionals without classes. Policymakers must establish accountability and feedback mechanisms, supervise ethical standards for AI users, and prevent vulnerable enterprises from being bullied or exploited by erroneous data, left without income or recourse, while negligent parties continue to enjoy high salaries—ultimately resulting in injustice.
In the AI era, students’ clarity of thought, judgment, and lifelong learning skills will determine whether they can truly grasp the logic behind technology, rather than merely using tools. This event is not only an academic advancement in AI but also a profound lesson for education: cultivating mathematics, logical reasoning, and comprehensive abilities is key to meeting the challenges of the future.
