2026年9月，中国发布第八批新职业共34个，其中11个为新职业、23个为新工种，数字职业占比达到45.5%1。截至目前，中国已累计发布113个数字职业1。本批新增职业涵盖数字孪生工程技术人员、具身智能机器人应用技术员、工业产品数字建模师、运动数据分析师、低空物流员、智能体开发员等多个领域1。
与此同时，AI在数学领域的突破也引发了学术界的担忧。2026年9月11日，OpenAI的GPT-6 Astra模型攻克了FrontierMath Tier 4中唯一未被解出的难题1。FrontierMath Tier 4包含43道研究级高难度数学题，通常需要专业数学家用数小时甚至数天才能解决1。更早前的9月8日，该模型仅用88小时就完成了"维纳-斯托克斯存在性与光滑性问题"的完整证明1。
这些成就引发了数学界的集体反思。陶哲轩等25位数学家和菲尔兹奖得主联合发布声明《人工智能在数学中的严重错位》，指出AI能力的提升与数学科学发展的目标存在严重偏离1。
China unveiled its eighth batch of new occupations in September 2026, comprising 34 positions across emerging sectors.1 Among these roles, 45.5% are centered on AI-related digital professions, bringing the cumulative count of digital occupations recognized in China to 113.1 The newly designated positions include digital twin engineering technicians, embodied intelligent robot application specialists, industrial product digital modelers, sports data analysts, low-altitude logistics workers, and intelligent agent developers.1
The same month witnessed a significant milestone in artificial intelligence capabilities that prompted swift response from the global mathematics community. On September 11, OpenAI's GPT-6 Astra resolved the final unsolved problem in FrontierMath Tier 4, a benchmark containing 43 research-level mathematics questions that typically require professional mathematicians hours or even days to address.1 In an earlier demonstration of computational power, OpenAI's model generated a complete proof of the Navier-Stokes existence and smoothness problem in just 88 hours on September 8.1
The rapid advancement sparked formal concerns among leading mathematicians. Twenty-five mathematicians and Fields Medal recipients, including Terence Tao, issued a statement titled "Serious Misalignment in Artificial Intelligence in Mathematics," expressing that AI capability development diverges significantly from the actual needs and directions of mathematical science.1
评论
还没有评论，欢迎留下第一条。