Claude Opus 4.7 is 20 times faster than humans: AI agent reprogrammed a robot dog without engineers' help
Anthropic continues to surprise the market. As part of the updated Project Fetch experiment, the Claude Opus 4.5 model fully took control of a four-legged robot — and did it 20 times faster than a team of human engineers. This is not just another benchmark: it is a demonstration of how an AI agent is beginning to displace humans from the cycle of development and configuration of physical systems.
In the first phase of the test, back in August 2024, participants with no robotics experience relied on AI only as an assistant. Now the scenario has changed dramatically. Claude Opus 4.5 worked almost autonomously under minimal researcher supervision. The neural network independently completed the entire cycle 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 results are impressive: Opus 4.5 turned out to be 18 times faster than a team using previous versions of AI, and 37 times faster than people working without chatbot assistance. The code written by the model turned out to be ten times more compact than human-written code. This speaks not only to speed, but also to a qualitatively different level of optimization.
Notably, Anthropic did not implement specialized algorithms for controlling the hardware. According to the team, the progress in robotics became a side effect of the general scaling of language models. This is an important signal for investors: universal AI systems are getting closer to working with the physical world as effectively as with code.
However, there were limitations. Claude still struggles with tasks requiring precise real-time physical feedback. The model guided the robot to the goal, but could not gently push a ball — a task where humans remain unmatched for now.
Anthropic is confident that the industry is entering an era of "physical AI agents." In the coming years, neural networks will be able to use standard tools and equipment as naturally as they work with software code today.
Expert opinion: This experiment is a clear signal for the market. The trend toward autonomous AI agents capable not only of writing code but also of controlling real devices will only gain momentum. Investors should take a closer look at projects integrating LLMs with robotics — this could become the next big growth driver in the crypto and technology sectors.