This article exposing the robotics industry is dropping later, and wow, it really dares to spill the tea. In a manufacturing plant, a robot extends its claw, picks up a bearing, turns around, and places it in a plastic box. A human could do this in 2 seconds, but the robot takes 70 seconds—and yet, this robotics company is valued at $1 billion. Mainstream general-purpose humanoid robots on the market cost an average of 200,000 yuan to produce, depreciate in less than 10 months, and need repairs after just 10 days of operation. Even slightly more advanced warehouse robots can move goods, but if the size or position of the goods is even slightly off, they fail. Most robotics companies still haven’t solved issues like stability and frequent human intervention. Using robots for production generally results in negative EV. The only business model that’s really working right now is robotics companies selling robots to data collection factories to generate sales revenue. The data factories then hire people to remotely control the robots, collect operational data, and sell it back to the robotics companies. Essentially, robotics companies are buying back part of the machine cost through data purchases. Meanwhile, local governments provide funding and land to support the data collection factories because these projects create jobs and come wrapped in the high-tech buzzwords of robotics and AI. This whole setup inflates valuations and paves the way for IPOs. Within the industry, it’s no longer about starting from scratch. Most startups are run by serial entrepreneurs pulling the strings behind the scenes. They’ll bring in a renowned industry expert or veteran to act as the front-facing CEO for fundraising. These guys really know how to play investors.