Experimenting

We are attempting to see what would happen if we were to integrate a Glyrin model into the Glyrin interface to filter and observe Glyrin's responses and actions to reduce the amount of a human monitoring the Glyrin models as much as possible. This will not only enhance safeguards, but we are also experimenting with this model to see if it can collect training data from the chats while filtering the training data that may be sensitive to the user. This will greatly speed up the collecting process of data for LLMs and also create a major boost in the privacy of the user. Of course, when the first Glyrin model is released and this experiment is a success, than we will give the user the option of the model collecting data or not, like most services do.