As we have moe models where it picks two of the expert models, can we apply this to conversations?
You have a history of conversations with descriptive titles, couldn't the ai pick relevant conversations as memory? Or make or change conversations at will to the appropriate conversation creating a sudo rag?
As we have moe models where it picks two of the expert models, can we apply this to conversations?
You have a history of conversations with descriptive titles, couldn't the ai pick relevant conversations as memory? Or make or change conversations at will to the appropriate conversation creating a sudo rag?
Add: hidden conversations, internal thought/dialogues.