Claude outperformed humans in controlling a robot dog: AI works 20 times faster than engineers
The world of robotics is undergoing a tectonic shift. Anthropic has unveiled the results of the second phase of the Project Fetch experiment, and the numbers are impressive. The Claude Opus 4.7 model demonstrated the ability to perform tasks for configuring and controlling a robot dog 20 times faster than teams of experienced human engineers.
While in August 2024, artificial intelligence only assisted employees without specialized experience, helping them find solutions faster, the situation has now changed dramatically. In the new testing phase, Claude Opus 4.7 operated almost entirely 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 robot control;
- created a motion trajectory monitoring system;
- configured an object recognition algorithm.
The test results are staggering. The Opus 4.7 model proved to be 18 times more productive than a team using previous AI versions, and 37 times faster than people working without chatbot assistance. But the key advantage is not just speed. The code generated by the neural network turned out to be 10 times more compact than that of human teams, indicating higher efficiency and cleaner execution.
A Side Effect of Scaling
It is important to note that Anthropic did not introduce specialized algorithms for controlling hardware. This progress became a side effect of the general scaling of language models. In other words, improving the AI's ability to understand and generate text automatically led to an increase in its competencies in the physical world.
However, a complete victory has not yet occurred. Claude still struggles with precise physical manipulations. The model successfully guided the robot to the target but could not gently push a ball to the right spot. This task requires complex real-time feedback, where humans still maintain superiority.
The Era of Physical AI Agents
Anthropic is confident: the industry is entering the era of "physical AI agents." In the near future, neural networks will be able to use standard tools and equipment as effectively as they work with software code today.
Recall that on June 13, Anthropic suspended access to the Fable 5 and Mythos 5 models due to a directive from the U.S. government under export controls. This underscores how seriously regulators view the potential of modern AI systems.
Expert Opinion: The results of Project Fetch are not just a laboratory curiosity but a clear signal to the market. We are witnessing a convergence of language models and robotics that will radically transform logistics, manufacturing, and the service sector in the next 2-3 years. Investors should closely monitor companies that integrate AI into physical systems, not just software.