文章日期为2026年2月12日（净资产截至2月2日），描述一批在美中科技脱钩与北京推动科技自主之际崛起、在中国发展的AI创业者，正累积可观财富。MiniMax Group的闫俊杰（Yan Junjie）表示，4年前投资人曾把他的多模态AI愿景斥为骗局，但这家上海公司如今估值达数十亿美元规模，并让他在36岁时成为亿万富豪；他也以OpenAI为对标来评测模型。与DeepSeek创办人梁文锋（Liang Wenfeng）和Unitree Robotics创办人王兴兴（Wang Xingxing）一起，一群多为男性、出生于1970年代与1980年代的创业者正在挑战美国在AI领域的主导地位，同时比上一代名人型高管更低调（对比Jack Ma于2017年阿里巴巴周年庆的表演）。

整体而言，这些中国AI富豪的净资产合计估计为$100.5 billion，与Bill Gates的$105 billion大致相当，但远低于Nvidia CEO Jensen Huang的$153 billion，凸显其规模同时也显示与美国AI硬体繁荣的最大受益者仍有差距。随著中国AI相关企业推进IPO，名单持续扩大，包括MetaX Integrated Circuits Shanghai Co.的陈维良（Chen Weiliang）与Moore Threads Technology Co.的张建中（Zhang Jianzhong）；据报导在12月的一波IPO中，约新增半打亿万富豪，其中许多人曾是Nvidia与Advanced Micro Devices的员工。曾任Nvidia高管的张建中用5年把Moore Threads打造为一家$45 billion公司，定位为替代他过去销售的那类晶片；而曾任AMD资深总监的陈维良在制定MetaX计划时，还招募多位AMD同事进入管理层。

一个关键特征是把低调当作风险管理：创办人力求避免被列入美国制裁/实体清单，并降低在国内因炫耀性财富而受到的审视；有观察者称美中关系的「toxic」状态是原因之一。教育背景与与国家相关的履历反复出现，例如寒武纪（Cambricon）的陈天石（Chen Tianshi）16岁上大学、在中国科学院工作6年，其财富自2024年初以来上升超过800%至$21.5 billion，与「buy-local」要求相关。Bloomberg的方法论聚焦于以AI基础设施、应用或硬体为主要业务且占多数营收来源的公司，排除多元化综合企业（如Alibaba与Tencent），并指出私募估值存在不确定性；它也标示梁文锋的净资产可能出现极端差异，从约$1 billion到超过$150 billion，反映在没有风投历史的私人公司定价上的困难。

Dated February 12, 2026 (with net worths as of Feb. 2), the article describes a new cohort of China-based AI founders building large fortunes amid US-China tech decoupling and Beijing’s push for technological independence. MiniMax Group’s Yan Junjie says that 4 years earlier investors dismissed his multi-modal AI vision as a scam, but the Shanghai firm is now valued in the multibillion-dollar range and made him a billionaire at age 36, as he benchmarks models against OpenAI. Alongside DeepSeek founder Liang Wenfeng and Unitree Robotics founder Wang Xingxing, a growing set of mostly male founders born in the 1970s and 1980s are challenging US AI dominance while keeping a lower public profile than prior-era celebrity executives (contrasted with Jack Ma’s 2017 Alibaba anniversary performance).

Collectively, these China AI tycoons are estimated to hold $100.5 billion, roughly on par with Bill Gates’s $105 billion but far below Nvidia CEO Jensen Huang’s $153 billion, underscoring both the scale and the remaining gap versus top US beneficiaries of the AI hardware boom. The list expands as Chinese AI-related firms pursue IPOs, including MetaX Integrated Circuits Shanghai Co.’s Chen Weiliang and Moore Threads Technology Co.’s Zhang Jianzhong; in December, a cluster of IPOs reportedly minted roughly half a dozen new billionaires, many of them former employees of Nvidia and Advanced Micro Devices. Zhang, an ex-Nvidia executive, built Moore Threads over 5 years into a $45 billion company positioned to replace the kinds of chips he once sold, while Chen, a former AMD senior director, developed MetaX’s plan while recruiting multiple AMD colleagues into management roles.

A defining pattern is discretion as risk management: founders aim to avoid placement on US sanctions/entity lists and to reduce domestic scrutiny of conspicuous wealth, with one observer citing the “toxic” state of US-China relations as a driver. Educational and state-adjacent pedigrees recur, such as Cambricon’s Chen Tianshi, who entered college at 16, worked 6 years at the Chinese Academy of Sciences, and whose wealth rose more than 800% since the start of 2024 to $21.5 billion, tied to “buy-local” mandates. Bloomberg’s methodology focuses on firms where AI infrastructure, applications, or hardware are the primary business and majority revenue source, excludes diversified conglomerates (e.g., Alibaba and Tencent), and notes uncertainty in private valuations; it flags an extreme variance for Liang Wenfeng’s net worth from about $1 billion to more than $150 billion, reflecting the difficulty of pricing a private company with no venture capital history.