Summary of Key Points
The WAIC 2026 robotics exhibition seemed incredibly lively: hundreds of intelligent furniture manufacturers gathered, and robots were capable of dancing, doing acrobatics, and demonstrating various scenarios. However, the author saw through the surface to underlying concerns—most of these robots were products of “assembly factories” lacking genuine technological innovation. The demonstration scenarios were all pre-programmed and failed in different environments. The commercialization path was unclear, driven by capital rather than real demand, reminiscent of the past craze for Clivia plants, which could potentially turn into a bubble-like memory known as a “Chinese dream core.”
I. Surface Festivity: The Robotics “Township Amusement Park”
The scene at the exhibition resembled the old-fashioned amusement parks in rural areas during my childhood: parents brought their children to watch robots dance, distribute medication, or act endearingly, while manufacturers handed out business cards, and attractive representatives asked whether investors should invest in stage one or stage two projects. Hundreds of intelligent furniture manufacturers (many of which had not even been established the previous year) showcased a wide range of products. For example, Zhiyuan’s Expedition A3 Ultra was the “highlight of the exhibition,” Yushu’s humanoid mecha attracted attention, and Fourier Robotics’ robots could fold clothes. Yet, all this excitement was like the colorful lights and children’s songs in an amusement park—appealing but somewhat absurd. Everyone was showing off what the robots could do, yet no one clearly explained what they should actually be capable of.
II. Hollow Core: All “Assembly Factories,” No Real Breakthroughs
The seemingly diverse range of robots actually represented little more than superficial changes:
- Concentrated Areas of Innovation: Most companies focused on three main areas: humanoid robots, quadruped robots, and robotic arms. Innovation was limited to basic metrics such as “doing acrobatics vs. dancing” or “lifting 300 pounds vs. lifting 500 pounds.”
- Dependence on Upstream Suppliers: Core components like reducers and sensors were supplied by a few key suppliers, making these companies essentially “assembly factories.” (This is similar to smartphone assembly, but smartphones have a real demand; robots do not.)
- Lack of Fundamental Innovation: The so-called “embodied intelligence” still relied on large models and robotic arms—the large models handled conversations, while the robotic arms performed pre-programmed actions. The actual “world model” that would enable robots to understand the real physical world remained merely theoretical, presented in PowerPoint presentations.
III. Pretended Demonstrations: Like “Rehearsed Public Lessons”
The smooth movements of the robots on the exhibition stage were like a well-rehearsed script:
- Pre-programmed Scenarios: Demos such as pharmacy sorting, factory palletizing, and screw tightening followed fixed procedures. These tasks failed when the environment changed (e.g., with additional obstacles or different target objects). The author witnessed manufacturers’ employees calling for help, saying, “The environment is different; the demonstration didn’t work!”
- The Need for Performance: Why perform these demonstrations? Because “industrialization” is the current political orthodoxy. Without showing practical applications, companies couldn’t secure capital funding. It’s like public lessons in school where teachers rehearse repeatedly, and students wear white shirts, smile, and raise their hands, yet they learn nothing substantial.
IV. Commercialization Dilemmas: A Clivia-Style Bubble
The current robotics craze echoes the 1980s’ fever for Clivia plants:
- Meager Profit Potential: Although over twenty manufacturers claimed to have industrial applications, and a dozen mentioned retail or home use cases, only a few are truly profitable on a large scale. Most projects are free trials without any paid orders.
- Poor Cost-Effectiveness: In industrial settings, AGVs (automated guided vehicles) or robotic arms are more suitable for warehouse sorting; in domestic services, robots are costly and unable to handle complex tasks, making them less effective than hiring hourly workers.
- Driven by Capital: The main topic of conversation at the exhibition was not about buying robots but which ones to invest in. Embodied intelligence has become a cover story for capital-seeking companies; traditional AI firms are just taking advantage of the hype to raise money rather than addressing real needs.
V. Lessons from Japan’s Robotics Boom
In 1980, Japan declared that year as the “Year of Robot Popularization.” With policy support, its robot inventory quickly surpassed that of the United States. However, without fundamental technological breakthroughs, the bubble eventually burst. Today, we face similar issues:
- Excess Production: This year, the production of humanoid robots is expected to exceed 100,000 units, but where will these robots work? How much value will they create? No one can answer these questions.
- Risk of Being Abandoned: One day, abandoned warehouses may be filled with rusting robots, like faded plastic dolls that were once dazzling but became symbols of a failed technology bubble.
Behind the excitement, we must ask: Are robots tools designed to solve problems, or are they merely toys for capital speculation?
(The entire analysis uses everyday metaphors such as “amusement parks,” “public lessons,” and “Clivia plants” to explain complex financial and technological issues in a way that non-financial readers can understand the risks associated with the robotics industry.)