The AI model Claude Opus 4.7 outperformed humans by 20 times in controlling a robodog: a new stage in the evolution of autonomous agents
The world of robotics and artificial intelligence has made another qualitative leap. As part of the updated Project Fetch experiment, Anthropic demonstrated that its flagship model, Claude Opus 4.7, can perform complex tasks of configuring and controlling a four-legged robot 20 times faster than teams of human engineers. This is not just another benchmark — it is a signal that the boundaries between software code and the physical world are blurring at an incredible speed.
For context: in August 2024, company employees with no robotics experience attempted to program a robot dog using AI. At that time, the neural network served only as an assistant, speeding up the search for solutions. Today, the situation has changed dramatically. Claude Opus 4.7 worked almost autonomously, under minimal researcher supervision. The neural network independently completed a whole range of tasks:
- Connected to video sensors and LiDAR;
- Wrote a program for manual control;
- Created a trajectory monitoring system;
- Configured an object recognition algorithm.
The numbers are impressive: the Opus 4.7 model proved to be 18 times faster than a team using previous AI versions, and 37 times faster than humans working without chatbot assistance. Moreover, the code generated by the neural network turned out to be 10 times more compact and efficient than human-written code.
The key conclusion drawn by the developers: progress in robotics is a side effect of the general scaling of language models. Anthropic did not introduce specialized algorithms for controlling hardware — this is pure magic of learning from vast data sets.
However, not everything is so smooth. Despite the triumph in programming and navigation, Claude still faces serious difficulties with precise physical manipulations. The model successfully guided the robot to its target but could not gently push a ball to the desired point. This task requires complex real-time feedback, where humans still hold the advantage.
Anthropic believes the industry is entering an era of "physical AI agents." In the near future, neural networks will be able to use standard tools and equipment as effectively as they currently work with code. Notably, on June 13, the company suspended access to the Fable 5 and Mythos 5 models due to a directive from the U.S. government on export controls — this underscores the growing strategic importance of such developments.
Expert opinion: The current experiment is a clear indicator that we are transitioning from AI as a "tool for thinking" to AI as a "tool for action." The key challenge now is not in writing code, but in fine motor skills. Once models learn to manage physical feedback with the same efficiency as digital feedback, we will witness a true boom in autonomous robotics.