Simple encoding is, of course, effective, but it is tied to a specific library implementation and cannot be read by a person without decoding.
However, approximately the same effect can be achieved by using any of the regular (artificial) languages, for example, Esperanto, as inter-AI communication.
Their word formation rules are strictly formalized, due to which encoding and decoding occurs very quickly using conventional algorithmic methods (without using neural networks).
But this method of communication has a very big advantage due to the usual human readability and a potentially larger number of assistants in the form of enthusiasts of this language.
Simple encoding is, of course, effective, but it is tied to a specific library implementation and cannot be read by a person without decoding.
However, approximately the same effect can be achieved by using any of the regular (artificial) languages, for example, Esperanto, as inter-AI communication.
Their word formation rules are strictly formalized, due to which encoding and decoding occurs very quickly using conventional algorithmic methods (without using neural networks).
But this method of communication has a very big advantage due to the usual human readability and a potentially larger number of assistants in the form of enthusiasts of this language.