@46ge5: Let me tell you a little secret. A certain scam robot company stopped investing heavily in R&D because scaling robot algorithms requires extremely expensive data collection and training. They already know they have no hope of creating the GPT of robotics (they've tried various VLAs, world models, and RL, all to no avail).
Summary
An anonymous tip reveals that a robot company has abandoned R&D due to the high cost of algorithm scaling and the failure of VLA, world model, and RL approaches. Instead, they are now building large robot toys to fool investors.
View Cached Full Text
Cached at: 06/08/26, 09:32 PM
Let me tell you a little secret: a certain scam robot company stopped investing heavily in research because the data collection and training required to scale robot algorithms are extremely expensive. They already clearly know they have no hope of creating the GPT of the robotics world (they’ve tried all sorts of VLA, world models, and RL, and none of it worked).
So they chose to build large-scale Transformers toys to fool the naive.
Similar Articles
@FinanceYF5: Chinese robot companies are still in their Series A rounds, yet they've started investing everywhere like VCs. 29 embodied intelligence companies have made a cumulative 125 external investments. AgiBot alone has invested in 37. The industry chain is using capital to lock in technology, teams, and positions in advance.
China's 29 embodied intelligence robot companies have made 125 external investments in their early financing stages, with AgiBot alone investing in 37 projects, using capital to lock in technology, teams, and market positions in advance.
@FinanceYF5: Cost recipe for humanoid robots: Linear actuators ~30%, Rotary actuators ~25%, Dexterous hands ~20%, AI + vision ~15%, Battery + thermal management <10%. Actuators + hands account for ~75% of the total BOM. High-precision hardware is still the biggest cost.
The tweet breaks down the cost structure of humanoid robots: linear actuators ~30%, rotary actuators ~25%, dexterous hands ~20%. Actuators and hands together account for ~75% of the total BOM, with high-precision hardware still being the largest cost component.
@seclink: Real Robot Interaction Data Is the Key Bottleneck for VLA Deployment: Model architectures are converging quickly, but generalizing to contact-rich long-horizon tasks such as warehouse picking, factory assembly, and home services still depends on large-scale diverse real-world data to improve success rates and throughput. A few companies have thousands to tens of thousands of hours of proprietary data (e.g., Physical Intelligenc…
The article points out that real robot interaction data is the key bottleneck for VLA deployment. Model architectures are converging, but data acquisition is difficult. A few companies own proprietary data that forms a moat, while open-source datasets such as Open X-Embodiment and DROID are available for reference and validation.
@lidangzzz: 10 Questions to Identify Modern Tech Idiots: 1. Believing that the Unitree Robotics robots at the Spring Festival Gala have an AI model inside; 2. Believing that Chinese large model companies (deepseek, Alibaba, the Six Small Tigers) selling API is actually selling electricity, a crushing dividend of industrial Cthulhu + infrastructure mania — the US is so short on electricity now that OpenAI can't even afford lights…
This article lists ten common misconceptions about the Chinese tech and AI industry, satirizing the exaggerated belief that Chinese companies are surpassing the US.
@0xCheshire: Chamath just revealed a deeply unsettling truth for the AI industry. He asked his own CTO to review the company's spending, and the result was staggering: "Our token costs are doubling every 45 days," yet the downstream productivity gains are at most about 5%. Costs are skyrocketing exponentially, while returns remain basically flat...
Chamath reveals the harsh reality of AI costs vs. returns: token costs double every 45 days, but downstream productivity gains are at most 5%. Large model capability improvement has hit an asymptote, and within the next 3-4 years, every company will face an ultimate reckoning between cost and benefit.