Nope.

The article clearly spells out that it's end to end LLM. Text and video in, motor function out.

Technically, the text model probably has a few copies, but they are nothing more than Asimov's narrative. Laws don't (and can't) exist in a model