Anthropic tested Claude Fable 5.2 in electronics design

Anthropic has tested the Claude Fable 5.2 model in the field of real-world hardware and electronics design. During the experiment, the AI participated in creating three functional device prototypes: a speakerphone, a voice assistant hub, and a virtual pet.

In tests on the PCB Gym benchmarks, the system achieved a result of 69%. In addition to calculations and printed circuit board design, the model successfully coordinated the work of human engineers, providing step-by-step instructions for assembling and debugging actual hardware.

The results obtained expand the practical application of language models beyond writing software code into physical engineering. This allows engineers to use AI as a full-fledged technical assistant when creating complex hardware systems.