Claude Opus 4.7 crushed humans in robot dog control: 20 times faster
A new phase of Anthropic's Project Fetch experiment has shown a stunning breakthrough: the Claude Opus 4.7 language model handled the setup and control of a four-legged robot 20 times faster than teams of human engineers. This is not just an algorithm's victory—it's a paradigm shift in robotics.
Let me recall the context: in August 2024, company employees with no robotics experience tried to program a robot dog using previous versions of AI. At that time, the neural network only acted as an assistant. Today, everything is different. Claude Opus 4.7 worked almost autonomously, under minimal researcher supervision. The model independently:
- connected to video sensors and LiDAR;
- wrote a program for manual control;
- developed a trajectory monitoring system;
- configured an object recognition algorithm.
The numbers speak for themselves: Opus 4.7 was 18 times faster than the team using older AI versions, and 37 times faster than people working without a chatbot. Moreover, the volume of generated code was 10 times smaller than that of human teams—the neural network wrote not only faster, but also more efficiently.
An important nuance: the experiment's authors emphasize that this progress is a side effect of the general scaling of language models. Anthropic did not introduce specialized algorithms for controlling hardware. This means that with each new generation, AI will automatically acquire skills for working with physical objects.
However, not everything is so smooth. Claude still faces serious difficulties with precise physical actions in real time. The model managed to guide the robot to the target, but failed the task of gently pushing a ball to a specific point. Such operations require complex feedback, where humans still maintain an advantage.
At Anthropic, they are confident: we are entering the era of "physical AI agents." In the near future, neural networks will be able to use standard tools and equipment as naturally as they currently work with software code.
My expert assessment: This experiment is a signal for the entire industry. If language models continue to scale at current rates, we will see AI not only writing code but also assembling, configuring, and repairing equipment. Investors, take a closer look at companies integrating LLMs into hardware—this is the next big trend.