Back in 2020, I was sitting in my office at Danone in Shanghai with the China marketing team. We were deep into consumer engagement models, planning our influencer marketing campaign for Aptamil. In this kind of work, you spend weeks negotiating with KOLs, briefing them on product messaging, reviewing content drafts, and crossing your fingers that the final video would feel authentic. It was exciting work. But it was also exhausting. We knew that consumers trusted real people

The Moment Everything Changed In early 2026, a robot arm with a table tennis paddle did something unprecedented: it defeated a former world No. 5 player in a regulation match. The robot had never trained in the real world. It learned everything in simulation - and then stepped onto a real court and won. This wasn't a lab demo. It was a signal. Sony AI's Project Ace bridged the gap between virtual intelligence and physical execution. The same reinforcement learning "brain" tha

Yesterday, the news of UBTECH's U1 Series humanoid robots caught my attention. Priced from 119,800 yuan to 990,000 yuan. Over 13,361 pre-orders already been placed. A 2-4 hour battery life. Can't walk (except the top-tier version). Can't cook. Can't do household chores. Yet people are buying. It reminded me of how popular Jellycat has become. A $70 plush bunny. People queue for 1-2 hours just to watch staff "whip cream" on a stuffed dessert in London, Paris, and Shanghai pop-

Most organizations believe they are undergoing digital transformation. In reality, the vast majority are simply digitizing - building apps, automating processes, and optimizing existing operations. These are necessary steps. But they are not transformation. Transformation happens when you use digital to redefine your business model itself. I’ve been exploring this question with executive teams across industries, most recently at Walmart China’s leadership programs. One framew

The year 2026 marks a critical inflection point for the humanoid robotics industry. Global shipments are projected to surge from 15,000 - 20,000 units in 2025 to over 50,000 units in 2026. But the real story isn't the numbers themselves; it's that the race is shifting from "can it walk" to "can it learn." Morgan Stanley's latest industry report states that the "robot brain" technology architecture remains unsettled, and competitive moats are increasingly tilting toward those

On April 27, 2026. An AI agent startup got blocked from selling to foreign buyers. The reason wasn't valuation. It was data governance. Image credit: https://www.bbc.com/news/articles/ce3k11q9qe1o The Trigger In April 2026, China's National Development and Reform Commission (NDRC) – through its investment arm – issued a prohibition notice against a foreign acquisition of Manus, an AI agent startup building the orchestration layer between large language models and real‑world b

Overview On April 19, 2026, Beijing Yizhuang hosted the world’s first half‑marathon for humanoid robots. This was not a gimmick but a rigorous engineering test: 21.0975 km, 22 turns, slopes, and outdoor conditions. EastGate AI analyzed the results to assess locomotion, autonomy, and durability. Key Results Participation growth The second edition of the event saw substantial growth. In 2025, only 20 teams participated, with roughly six brands or robot types represented and no

Overview In mid‑April 2026, EastGate AI accompanied a group of European operations leaders on a guided tour of China Humanoid Robotic Ecosystem Conference. Our objective: assess real industrial capabilities, benchmark performance claims, and identify commercially viable use cases. What we observed confirms a clear gap between marketing narratives and on‑the‑ground productivity. Key Observations 1. Diverging strategic focuses Unitree continued to showcase dynamic, entertainmen

我们的创始人 Mabel Lu 本周在中国数字化创新博览会上分享了一个核心观点：智能悖论——慢AI，快增长。为什么在AI上“慢下来”，反而能带来更快的增长和更深的客户关系。 这个悖论很简单，但很有力量： 在“人”的事情上要快——信任、共情、连接。在“替代人”的事情上要慢——甚至刻意慢。 当你把一段关系中的人性优化掉，你也把这段关系的价值优化掉了。 核心原则 在展开案例之前，先记住一个基本事实： AI 处理事务性工作，人类主导关系性工作。 事务性 – 速度、规模、效率。交给 AI。 关系性 – 信任、情感、连接。留给人类。 把这个分清楚，大部分 AI 的问题就开始消失了。 三个案例 1️⃣ 速度陷阱——某大型公用事业公司 一家公用事业公司快速部署了 AI 客服。两个月上线。 成本降了，效率指标也好看。 但 AI 听不懂方言，读不懂情绪，转人工时没有任何上下文。 呼叫量没有下降，反而上升了。人工花在“修复 AI 错误”上的时间，比自己做还多。 问题不是 AI。问题是忘了为“中间的人”设计。 2️⃣ 社交电商飞轮——一位12岁的小创业者（创始人女儿 ���

Our founder, Mabel LU, spoke at China Digital Innovation Expo (CDIE) in Shanghai this week about the Intelligence Paradox: slow AI, faster growth. Why slowing down on AI creates faster growth and deeper customer relationships. The paradox is simple but powerful: Move fast on what makes you human – trust, empathy, connection. Move deliberately – even slowly – on replacing humans. The Core Principle Before we dive into the cases, one foundational truth: AI handles the transacti

China’s humanoid robot industry has reached a critical inflection point. According to the latest industry data, more than 140 companies are now active across the supply chain—ranging from core components to full‑system integration and application deployment. This makes China one of the most dynamic and fast‑evolving humanoid robotics markets globally. For international observers, the scale and speed of this buildup matter. China is not only a major manufacturing base but als

中国人形机器人产业正迎来关键转折点。根据最新行业数据，目前已有 超过140家企业 活跃于整条供应链 ---从核心零部件到系统集成，再到应用落地。这使得中国成为全球最具活力、发展最迅速的人形机器人市场之一。 对国际观察者而言，这一产业集聚的规模和速度具有重要意义。中国不仅是重要的制造基地，在驱动器、控制系统和具身智能领域也正成为创新的重要来源。了解这一生态的构成与发展脉络，对于任何关注机器人未来走向的人而言都至关重要。 三大类玩家 这140余家企业可分为三个层次，各具优势。 1. 科技巨头与资本雄厚的独角兽 华为、比亚迪、宁德时代等企业依托其在人工智能、精密制造、动力系统等领域的技术积累，为人形机器人提供底层支撑。与此同时，宇树科技、智元机器人、星动纪元、逐际动力等一批获得资本重注的初创公司快速从原型机迈向实地测试。其中多家企业在双足运动控制、动态稳定性及非结构化环境下的操作能力方面已展现出令人瞩目的成果。 2. 工业机器人领域的老牌企业 埃斯顿、新松机器人、埃夫特等传统自动化企业正将其工业机器人领域的经验迁移至人形平台，重点聚焦可靠性、成本控制与

By 2035, AI will evolve from a digital tool into a physical workforce of 900 billion autonomous agents - creating trillion-dollar markets in healthcare, mobility, and manufacturing while demanding infrastructure investments that will reshape global energy and computing. Top 6 takeaways: 1. 🤖 The 900 Billion Agent Economy: 900B agents serving 9B people. Every customer gets an AI "co-pilot"; every employee gets an AI "teammate." Competition shifts from "best app" to "most trus

Over the weekend, I kept coming back to one question: If AI handles the logic, what's left for humans? The report offers a compelling answer that cuts through the "AI will take our jobs" noise. As AI evolves from a tool into a decision-making partner, the human role shifts from executor to value judge. Here's what that looks like across industries: 🏥 Healthcare (Page 67) AGI will handle the heavy lifting - integrating medical records, genomic data, and real-time monitoring t

Last weekend in Shanghai during the Appliance & Electronics World Expo (AWE) 2026. I tried Qianwen AI Glasses G1. As a mom who cooks a lot for my two teenagers, I had one question: Can this actually help in the kitchen? Three things surprised me: 1. It played my music. I asked for something to cook to, and it knew what I like. No scrolling. Just my vibe, instantly. 2. It handled translation effortlessly. I tested it with a recipe, and it read everything back clearly. No phone

Just wrapped up 2026 Appliance & Electronics World EXPO (AWE) in Shanghai. If CES is where the world dreams about the future, AWE is where that future lands in your living room. Here's how AI is finally moving from concept to reality. 1. Physical AI: from chatbots to helpers • MagicLab & Zeroth: humanoid robots as companions, elderly care, pet monitoring.
• Dreame: AI rings (ECG), lawn mowing robots (98% obstacle recognition), wheelchair robot with robotic arm, redefining eld

On the second day after visiting the robot exhibition, I went to GAIC – the Global AI Conference happening at the same time in Hangzhou. If the former showcased the evolution of "bodies" – agile robots with increasingly flexible joints and sophisticated tactile senses – the latter revealed the impact of "brains." A few moments from GAIC today left a deep impression on me: 🔹 "0-0-0, 1-1-1" - One CEO shared their hiring philosophy: hire only post 00s (born after 2000), zero e

I'm currently in Hangzhou, an amazing tech valley in China that has become a powerhouse for AI and robotics innovation. From watching these robots dazzle on the CNY Gala stage to seeing them up close in exhibition halls, the energy is palpable. What strikes me most is the clear differentiation among the leading players. Each has carved out a unique niche: • Unitree: Legged locomotion champion. Global leader in quadruped sales (70% market share). From backflipping H1 to agile

Have you started "raising lobster" yet? It's the new buzzword in AI circles. OpenClaw, an open-source AI agent, has exploded overnight. The hype, in numbers: · 13+ major Chinese tech companies launched integrations in weeks · Millions experimenting, with lines wrapping around buildings to get it (see pic from Tencent HQ 3 days ago) · Why? Because AI is finally moving from conversation to execution For marketers, this matters. For the first time, AI can truly act - not just ad

Watching the 2026 Spring Festival Gala, one thing was clear: robots are no longer just stage props. They were central, charismatic performers! What struck me most was the perfect blend of cold machinery with a warm festive spirit. It created a unique dialogue between our rich cultural traditions and modern innovation. And this wasn't just for show - the robots were incredibly flexible and high-performing, executing complex moves alongside human artists with zero errors. When