Claude AI learned to control a robot dog 20 times faster than humans: a new stage in the evolution of autonomous agents
The artificial intelligence industry is taking another decisive step into the physical world. As part of the updated Project Fetch experiment, Anthropic's Claude Opus 4.7 model demonstrated the ability to configure and control a four-legged robot at speeds 20 times faster than human engineering teams. This is not just a code test — it is a paradigm shift in AI's interaction with hardware.
Let me remind you that in August 2024, company employees with no robotics experience attempted to program a robot dog using AI. At that time, the neural network acted only as an assistant, accelerating the search for solutions. Today's reality is different: Claude Opus 4.7 worked almost completely autonomously, with minimal researcher supervision.
What did the neural network do?
The model independently completed a set of tasks that typically require a team of specialists several days:
- Connected to video sensors and LiDAR;
- Wrote a program for manual robot control;
- Created a trajectory monitoring system;
- Configured an object recognition algorithm.
The numbers speak for themselves: Opus 4.7 was 18 times faster than a team using previous AI versions, and 37 times faster than humans working without chatbot assistance. Moreover, the volume of code written by the neural network was 10 times smaller than that of human teams. This is direct evidence that modern models do not just repeat patterns but generate more efficient and concise solutions.
It is particularly noteworthy that Anthropic did not introduce specialized algorithms for controlling the hardware. The progress in robotics became a side effect of the general scaling of language models. This means that with each new generation, AI will automatically acquire skills for working with physical objects.
However, one should not rush to conclusions about the complete replacement of humans. Claude still struggles with precise physical manipulations. The model managed to guide the robot to its target but failed at the task of gently pushing a ball to a specific point. This requires complex real-time feedback — an area where humans still maintain superiority.
Anthropic rightly notes that 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 work with software code today. For the crypto industry and DeFi, this opens up enormous prospects — from automating mining farms to managing robotic warehouses for storing physical assets.
Expert opinion: While Claude failed at the "push the ball" task, reminding us that the physical world requires fine motor skills and adaptability that AI has not yet mastered. But if the pace of progress continues, within 2-3 years we will see the first fully autonomous robots in crypto warehouses and logistics. Investors should take a closer look at projects at the intersection of AI and robotics — this is the next big trend.