AI agent Claude outperformed humans in controlling a robot dog: 20 times faster
Anthropic has presented the results of the second phase of the Project Fetch experiment. The Claude Opus 4.7 model demonstrated impressive results: it completed tasks for configuring and controlling a robotic dog 20 times faster than teams of human engineers. This is not just a quantitative leap—it is a fundamentally new level of AI autonomy.
In August 2024, company employees with no experience in robotics attempted to program a four-legged robot. At that time, AI only acted as an assistant, speeding up the search for solutions. However, in the new testing phase, Claude Opus 4.7 worked almost completely autonomously, under minimal researcher supervision. The neural network independently:
- connected to video sensors and LiDAR;
- wrote a program for manual control;
- created a robot path monitoring system;
- configured an object recognition algorithm.
The model's performance proved impressive: Opus 4.7 was 18 times faster than the team using older AI versions, and 37 times faster than humans working without chatbot assistance. Moreover, the neural network wrote more efficient code—its volume turned out to be 10 times smaller than that of human teams. This indicates that AI is capable not only of accelerating processes but also of optimizing them at a fundamental level.
The experiment authors emphasize that progress in robotics has become a side effect of the general scaling of language models. Anthropic did not implement specialized algorithms for controlling hardware—this is pure emergence.
Despite the success, Claude still struggles with precise physical actions. The model managed to guide the robot to the target but failed at the task of gently pushing a ball to the right spot. This requires complex real-time feedback, in which humans still maintain superiority.
My Analysis and Forecast
Anthropic rightly claims that the industry is entering an era of "physical AI agents." However, current limitations in precise motor skills remind us that AI is still far from fully replacing humans in tasks requiring fine sensorimotor coordination. Nevertheless, the pace of progress is such that within the next 2-3 years, we may see commercial solutions where AI controls robots in warehouses and manufacturing with efficiency unattainable for humans. This will change the labor market in industry faster than many expect.