文｜富充 编辑｜苏建勋 清华大学电子工程系长聘教授汪玉 北京智源人工智能研究院院长王仲远 阶跃星辰创始人&CEO 姜大昕 星海图创始人&CEO 高继扬 原力灵机联合创始人&CEO 唐文斌 阶跃星辰创始人&CEO姜大昕首先提出"ChatGPT时刻"的定义标准，是"零样本泛化"——即使给出从未见过的指令，AI也能回答问题完成任务——这正是 大语言模型所展现的能力。 但姜大昕旋即指出，因为具身智能的泛化要涉及场景、任务、操作物体等更多维度，所以机器人要达到这个标准还十分困难。 作为机器人创企的CEO，高继扬进一步解释了具身智能商业化落地的难点：大语言模型可以"模型即产品"，终端是手机电脑、渠道是互联网传播；具身智 能却必须穿过更长的产业链——整机、供应链、真机数据、线下交付，缺一不可。 具身智能正在等待自己的"ChatGPT时刻"。但关于这个时刻的具体定义，业内还充满非共识。 近日，原力灵机的技术开放日圆桌论坛上，5位AI界的一线产、学、研从业者把这个问题摊开，各自发表了见解。他们分别是： 基于以上种种待解决的问题，原力灵机联合创始人&CEO 唐文斌，给出了一个眼下更可抵达的"具身智能ChatGPT时刻"：先在一个 ...

Fast Reading 在人工智能的诸多分支中，具身智能是近一年间被高频提及的方向之一。 从工业机器人到服务机器人，从自动驾驶到人形机器，各条技术路线每隔一段时间就会被寄予一次"通用智能入口"的期待。 但与算法驱动的软件革命不同，它始终被现实世界的摩擦力拖慢节奏。 如果只看公开视频，具身智能的公共叙事几乎被同一组类似的画面占据：机器人跑得更稳了，抓得更准了，动作更连贯，任务更复杂。融资在加速，模型 在迭代，具身智能似乎正处在一条确定无疑的上升曲线。 成功被循环播放，失败被剪掉——在实验室之外，另一种叙事同时存在：部署成本、稳定性、维护复杂度，仍在反复拉长商业化的时间表。 2月10日，原力灵机的首次技术开放日在北京中关村国家自主创新示范区展示中心举行。 原力灵机发布了三款核心产品：具身原生大模型 DM0、具身原生开发框架 Dexbotic 2.0、以及具身原生应用量产工作流 DFOL。这也是原力灵机成立近一 年来，公司核心团队第一次集体公开亮相。 在当天的"Physical Al Next 圆桌论坛"上，来自产、学、研的五位嘉宾花了约一半的时间，讨论一个问题： 具身智能的ChatGPT时刻何时到来？ 具身智能 ...

Core Insights - NVIDIA's CEO Jensen Huang announced that the robotics field has entered a "ChatGPT moment" during the CES 2026 keynote [1] - A series of open-source "physical AI" models were introduced at the event [1] Group 1 - Huang showcased a pair of charming BDX robots during his presentation [1] - The demonstration included how the robot "GR00T" learns to become a robot [1]

Core Insights - The humanoid robot industry is experiencing rapid product iteration, but achieving general capabilities akin to "ChatGPT" will take at least 2 to 3 years due to technological bottlenecks related to data accumulation and model training [1][4]. Group 1: Industry Trends - The humanoid robot applications are primarily focused on four areas: education, exhibition guidance, standard platforms for developers and researchers, and manufacturing/logistics, with the latter achieving success rates of 80-99.5% for specific tasks [2]. - There is a significant increase in consumer interest, as evidenced by higher-than-expected attendance from families and general consumers at the World Robot Conference (WRC), indicating strong short-term demand in consumer markets [2]. - Companies are launching more affordable and simplified products, such as the R1 robot priced at 39,900 yuan and the SA02 robot at 38,500 yuan, which may enhance market penetration [2]. Group 2: Product Development - Over 20 out of 50 humanoid robot exhibitors introduced new products in July and August, showcasing significant improvements in overall performance, speed, and fluidity compared to earlier in the year [3]. - The industry is still exploring transitional solutions to overcome challenges in training models effectively, particularly in complex environments that require high-quality real-world data [4]. Group 3: Technological Evolution - The best AI architecture for achieving general intelligence is still evolving, with discussions around the superiority of new models like Google's Veo3 over traditional VLA+RL approaches [5]. - The high domestic production rate of components, estimated at around 80%, is helping manufacturers control costs, although ongoing design optimizations are necessary to maintain competitiveness [5]. Group 4: Investment Insights - The report emphasizes the importance of focusing on segments of the supply chain with higher predictability, particularly in actuator assemblies, which are expected to have a clearer technological iteration path [6]. - The company Sanhua Intelligent Controls is rated as a "buy" due to its long-term growth potential in the humanoid robot actuator sector, with projected revenue and net profit CAGR of 19% from 2025 to 2030 [6].