AI models do not work that way. It is not input/output like traditional computing.
The model is actually generating the answer each time. The challenge comes in when we get into the weighting. That can reduce the accuracy. It is why more data is required, to create more connections within the vector database.
I didn't know the weighing concept. When you responded I sent your comment through Gpt4 to explain and now I've learned something new.