Crypto news

We are witnessing a historic breakthrough at the intersection of artificial intelligence and robotics. The Anthropic team has presented the results of the second phase of the Project Fetch experiment, and the numbers are impressive: the Claude Opus 4.7 model handled the setup and control of a four-legged robot 20 times faster than the best teams of human engineers working with the previous version of the AI.

Let me remind you that in August 2024, a similar test involved employees with no experience in robotics, and the AI only acted as an assistant, speeding up the search for solutions. Today's picture is fundamentally different. Claude Opus 4.7 operated almost autonomously, under minimal researcher supervision. The neural network independently completed a full cycle of tasks:

• connected to video sensors and lidar;
• wrote a program for manual robot control;
• created a trajectory monitoring system;
• configured an object recognition algorithm.

The key metric is speed. Compared to teams using older versions of AI, Opus 4.7 proved to be 18 times faster. And when compared to people working without chatbot assistance, the gap reaches 37 times. At the same time, the neural network generated code that is 10 times more compact than human code, indicating higher efficiency and a lower tendency for errors.

Particularly noteworthy is the fact that Anthropic did not implement specialized algorithms for controlling the hardware. The progress in robotics was a side effect of the general scaling of language models. This confirms the hypothesis that universal AI systems can adapt to physical tasks without additional training.

However, there were limitations. Despite success in navigation, Claude was unable to neatly push a ball to a specific point. Tasks requiring precise real-time feedback still remain the domain of humans. This indicates that for full-fledged physical interaction, AI needs additional sensory and motor algorithms.

At Anthropic, they are confident that we are entering the era of "physical AI agents." In the near future, neural networks will be able to control standard tools and equipment as naturally as they work with code today.

My analysis: This experiment is not just another benchmark. It demonstrates that AI is moving from the virtual space into the real world. For the crypto industry, this is especially important: autonomous agents capable of interacting with physical objects could radically change logistics, mining, and the management of decentralized physical infrastructure (DePIN). For now, humans retain an advantage in fine motor skills, but the gap is narrowing with each generation of models.