天网

China’s market structure improves in H1 as AI, robots drive emerging industries_我的网站

A |     播放          下一个          打开循环播放          00:00     /     00:00          倍速          3.0X     2.0X     1.5X     1.25X     1.0X     0.75X     0.5X          多音轨          AirPlay          0          静音播放中，点击 恢复音量          画中画          网页全屏          全屏          你可以 刷新 试试          视频信息     1.33.6          播放信息 上传日志 调试信息 [X]          视频ID     VID     -          播放流水     Flowid     -          播放内核     Kernel     -          显示器信息     Res     -          帧数     -          缓冲健康度     -          网络活动     net     -          视频分辨率     -          编码     Codec     -          mystery     mystery     -          按住画面移动小窗          21世纪经济报道记者 董静怡 上海报道          “AI六小虎”已走出截然不同的路线。    

    A Xinjiang local resident dances with a robot at the temporary exhibition hall at the region's first robotics 9S center on July 31, 2026. Photo: courtesy of the Xinjiang Future Era Information Technology Co, the operator of the project
China's market landscape experienced a structural optimization in the first half of 2026, powered by a lineup of emerging and future-oriented industries. 
Official data from the State Administration for Market Regulation (SAMR) shows 561,000 new enterprises were registered across China's strategic emerging and future industries, spanning next-generation information technology, high-end equipment manufacturing, generative artificial intelligence (AI) and humanoid robotics.
The rapid expansion of these new market players is revamping China's industrial layout and fast-tracking the cultivation of new quality productive forces, underpinning steady high-quality economic upgrading, experts said.
Cutting-edge technology sectors stood out as the brightest growth spots, posting explosive year-on-year expansion. The generative AI industry recorded a 28 percent surge in new start-ups with 55,000 newly established firms, while the humanoid robot sector saw a 9.5 percent rise, adding 116,000 new enterprises. 
"It shows that China's innovation momentum is exceptionally strong and remarkably broad-based — a trait that is virtually unique in the global landscape," Chen Jing, a vice president of the Technology and Strategy Research Institute, told the Global Times on Sunday. "The explosive growth of new enterprises in AI and robotics reflects the extraordinary depth and diversity of China's industrial ecosystem."
Outside high‑tech sectors, 3.957 million new services‑sector enterprises were established and booming emerging‑consumption businesses injected fresh impetus into domestic demand, the SAMR said.
The number of new firms in software and telecommunications services rose 3.3 percent to 322,000. Inspection‑and‑testing businesses soared 43.9 percent, and entities dedicated to the commercialization of scientific‑research achievements climbed 6.8 percent to 459,000.
Manufacturers accelerated industrial upgrading with companies moving up the value‑chain ladder, the SAMR said.
High‑end equipment industries have maintained growth during the first half, with 28,000 newly registered firms and a year‑on‑year increase of 2.5 percent. 
A total of 330,000 high‑tech‑manufacturing enterprises were operating as of the end of June, including 15,000 newly established ones. Start‑ups engaged in the production of spacecraft and launch vehicles rocketed 185.7 percent. New producers of fiber optic cables climbed 129.4 percent and integrated‑circuit businesses increased by 24.2 percent.
"We are witnessing a dual-engine growth model, where new consumption in the services sector and high-tech services reinforce each other, while the digital and real economies integrate at an accelerating pace. At the same time, the intelligent transformation of manufacturing is picking up speed," Chen said.
The dividends from this industrial transformation are generating a powerful multiplier effect, creating a virtuous cycle led by innovation, Chen noted.
"The first‑half growth and structural shifts among China's market‑based businesses mirrored the country's overall economic momentum and industrial makeup. New‑emerging sectors, core strategic pillar industries and forward‑looking future industries registered the fastest business expansion," Tian Yun, a veteran economist, told the Global Times on Sunday.
These solid achievements fully reflect the success of China's innovation-driven economic strategy.
Strong policy support for specialized, sophisticated, unique and innovative enterprises has fueled this industrial boom while these market players have helped China shore up, upgrade and extend its industrial chains, Tian noted.
This business boom lines up with this year's stellar foreign trade figures, Tian noted.
China's exports rose 23.9 percent year-on-year in dollar terms in July, data from the General Administration of Customs showed on Friday.
Exports of high-tech products, including industrial robots and 3D printers, surged by more than 50 percent year-on-year, outpacing the 39 percent growth recorded in the first half of the year, according to the Xinhua News Agency.
The growing pool of high‑quality market players signaled lasting improvements in China's global competitiveness, industrial upgrading capacity and the standard of goods and services. Broader product ranges, better‑made commodities and improved supporting services will deliver durable new‑growth momentum, Tian noted.
。         智谱AI和MiniMax相继上市，智谱创始人唐杰在内部信中表示未来两年不追求短期应用变现，将攻坚AGI；MiniMax创始人闫俊杰则表示不达到AGI将不领薪资。月之暗面同样将目标设定为“超越Anthropic等前沿公司，成为世界领先的AGI公司”。         另一边，阶跃星辰直接转向了AI硬件，发布了智能体手机；百川智能则选择在医疗垂直大模型上深耕；零一万物则一门心思扑向企业AI，走向了更细分的赛道。

B |          在WAIC 2026上，零一万物创始人兼CEO李开复向记者算了一笔账：中国目前有11家还在执着于做大模型训练的公司，扣除2家有特殊商业理由自研的，还有9家。         “如果全部是同质化的通用大模型，我觉得是不需要的。”李开复向记者表示，“除非真的是有信心或者是背水一战，天涯何处无芳草，为什么一定要选大模型？”          大背景是AI行业正在从“宏大叙事”阶段进入商业化阶段。高盛的一份报告指出，AI行业的终极赢家，将不再是那些投入资金最多的公司，而是那些在每美元AI投资中能产生最高“有用产出”的公司。         但目前，大部分AI公司都还是亏钱的。亏损的原因并不复杂。李开复分析可能有几个原因：一是训练模型很贵，即便有收入也可能难以覆盖成本；二是ToC公司买流量，流量成本也把利润消耗掉了。

C |          ToB被许多AI公司视为更有前景的方向。李开复也承认，从理论上说ToB的价值应该更大——普通用户不太可能每月花几千块钱去订阅一个Agent，而企业愿意为强化工作流或替代员工付费。但他同时坦承，中国ToB市场对项目制的依赖比较大。         所谓项目制，就是客户提出需求，AI公司按需定制开发，一单一议。这种模式下，要获得收入并不难，市场上对AI有好奇心和尝试意愿的企业很多，接单总能接到。但想要盈利，却不容易。         为了交付一个可用的产品，AI公司需要投入大量的人力进行定制开发、数据清洗、私有化部署，甚至提供驻场服务。这导致AI公司规模越大，亏损可能越严重，“做一单赔一单”。         行业普遍不赚钱的根源，还是在于技术价值尚未有效转化为商业价值。客户愿意为概念买单，但未必愿意为效果持续付费。这需要AI公司在产品定义、交付模式和客户选择上做出更清醒的判断。

Current article:http://06w08sg.cuntoujuefourantapai.cfd/3h94/ji84g.html

Published on:17:43:03