死寂

Is the rapid growth of China’s emerging industries driven by industrial subsidies?: People’s Daily_我的网站

格林第一季

一 |

Workers assemble solar photovoltaic modules at a smart manufacturing workshop of Ronma Solar Energy Group in Jindong district of Jinhua city, East China's Zhejiang Province, on July 28, 2026. Photo: VCG
China saw robust exports in electric vehicles (EV), lithium batteries and photovoltaic products, known as the "new three," in the first half of this year. More notably, robotics, artificial intelligence (AI) and innovative drugs, which represent the future direction of industrial development, are also emerging as new calling cards for China's foreign trade.
However, the impressive performance provoked unease among some Western media outlets and politicians. Some have deliberately portrayed China's rapid industrial development and strong competitiveness as a result of government subsidies, pushing the false claim that subsidies have created overcapacity and those low-priced Chinese products are flooding global markets. Such fallacies, which simply equate industrial subsidies with overcapacity, are not only logically flawed but also factually groundless.
In practice, many countries adopt industrial policies tailored to their national conditions and development needs, such as providing research and development (R&D) subsidies for emerging industries and risk related subsidies for agriculture.
Well-designed industrial subsidies can help address market failures, promote technological innovation and environmental protection, reduce poverty and support balanced development, rather than cause so called "overcapacity."
Multiple reports by the United Nations Conference on Trade and Development have noted that the number of industrial policies worldwide has grown rapidly over the past five years, with R&D subsidies, tax incentives and low interest loans for emerging industries becoming common international practices.
Forcibly linking industrial subsidies to "overcapacity" is, in essence, a political manipulation based on double standards. The US, for example, plans to provide $750 billion in various subsidies from 2022 to 2031 under its Inflation Reduction Act. Subsidized EVs are subject to requirements such as production and sales in the US or North America, effectively excluding other WTO members. US industrial subsidies for AI are even greater than those of all other countries combined.
Similarly, according to incomplete statistics, the European Commission is expected to provide more than 1.44 trillion euros ($210 billion) in various subsidies between 2021 and 2030. The EU's Industrial Accelerator Act links local content directly to financial support through "Made in EU" requirements, creating serious investment barriers and institutional discrimination.
Have these massive subsidies been labeled as causing "overcapacity"? The answer is no. While claiming that China's industrial subsidies lead to so-called overcapacity, these countries are themselves providing massive subsidies to their own industries. Such double standards amount to selective accusations targeting China, aimed at politicizing trade and economic issues and weaponizing industrial policy.
At a deeper level, accusations that "China's industrial subsidies cause overcapacity" are merely a pretext, reflecting growing anxiety and fear over the rising competitiveness of Chinese industries.
Looking back at the repeated hype in Western media, the criticism has consistently targeted China's most globally competitive industries, including new-energy vehicles, photovoltaics and power batteries. This exposes the real intention of shifting the blame for their own lagging industrial development onto China while stepping up restrictions against Chinese industries.
China's breakthroughs in these industries have been driven by advances in homegrown technologies, complete industrial and supply chains, and robust market competition, rather than by policy subsidies as some have claimed.
In recent years, China has taken multiple steps to regulate and improve its subsidy policies, from reviewing and correcting inappropriate local subsidies to exploring a unified negative list mechanism for local fiscal subsidies. China applies subsidies equally to all market entities, including foreign invested enterprises, strictly follows WTO rules, and continues to improve the compliance, effectiveness and transparency of its subsidy policies.
Rather than fabricating and hyping baseless claims about subsidies and obsessing over building trade barriers, certain Western media outlets and politicians should focus on addressing their own weaknesses and increasing investment in research and development. They should embrace healthy market competition with an inclusive mindset, promote mutual benefit through greater openness, and win markets and drive progress through genuine innovation.
This was compiled and translated by the Global Times English edition based on an article published in the "Chisu Jinsheng" economic commentary column of the People's Daily on August 10, 2026.
。

　阿里云优惠券先领券再下单 ERGO与ECODYNAMICS联合报告解析LLM的“内容审美” 结构清晰、问答模块化的内容，正成为AI驱动搜索时代的“新通行证”。近期，ERGO创新实验室与ECODYNAMICS联合发布的开创性研究报告在保险科技领域引发关注。这项覆盖33,000个AI搜索结果和600个网站的研究发现：大型语言模型（LLM）在呈现保险类内容时，显著偏好易读性强、结构良好且来源可信的信息——这一规律与传统搜索引擎优化（SEO）的核心原则高度重合。 核心发现：AI搜索与传统SEO策略的“不谋而合” 1. 内容结构化是“硬通货” 研究数据显示，采用模块化布局，尤其是问答形式的保险内容，被LLM（如ChatGPT）采纳生成答案的概率提升超40%。这种分段明确的组织形式便于AI提取关键信息，同时符合人类读者的认知习惯。

二 | 2. 可信度决定内容优先级 LLM在筛选信息时，会显著倾向标注清晰数据来源、作者背景及专业机构背书的内容。这与传统SEO中E-A-T原则（专业性、权威性、可信度）*完全一致。 3. 模型准确性差异显著研究对比了主流AI工具的可靠性：ChatGPT在保险类回答中的错误率接近10%，而专注垂直领域的you.com等平台错误率低50%以上。凸显专业领域需警惕“AI幻觉”风险。

三 | 行业启示：保险内容策略的转型方向 1. 从关键词堆砌到场景化问答保险企业需重构内容架构。例如，将“车险理赔流程”拆解为 “事故后5步操作指南”“如何在线提交照片证据”*等具体问题，适配LLM的答案生成逻辑。

四 | 2. 多模态内容提升权威感知研究指出，结合图文、图表或短视讯的解释性内容，能同步增强AI与用户的双重认可。例如健康险条款配疾病示意图，理赔指南嵌入流程图。 3. 专业大模型正在崛起针对通用LLM的局限性，行业已展开行动：如EXL公司近期推出保险专用大模型，通过领域微调使理赔数据解析准确率提升30%，成本降低30%。 未来趋势：AI搜索优化重塑保险服务链 本次研究印证了技术变革中的“不变法则”——内容价值始终居于核心。但AI时代的要求更为严苛： “LLM不是传统搜索引擎的替代者，而是进化者。它们迫使企业重新思考：如何用机器可读的方式，传递人类可信的信息。” ECODYNAMICS研究主管在报告中指出保险业应用已初见端倪：智能理赔机器人可解析用户上传的事故照片，自动对比保单条款；承保评估AI通过分析医疗报告影像，实现风险秒级判定；虚拟顾问**提供24小时保单解读，问答准确率依赖后端知识库的结构化水平。专家行动建议 1. 内容生产侧：建立“问答知识图谱”，将保险条款转化为层级化QA模块 2. 技术部署侧：接入行业专用LLM（如EXL保险模型），降低通用工具误判风险 3. 合规风控侧：对所有AI生成内容实施人工审核节点，尤其涉及赔偿金额与责任条款报告全文已收录于ERGO创新实验室2025年度《保险科技趋势白皮书》。