This is a hypothetical weapon and counterattack plan in case an artificial intelligence domination attempt would be deployed. The project is a single node, can be distributed and used freely. It could be replicated many times; its strength increases multifold with the number of nodes created. Artificial intelligence models are trained on data provided by humans. Text, image, movement, behavior, systems, structure ... many things can be replicated through enough input data. It is an open question whether intelligence, intention or even consciousness can emerge from the current attitude towards machine learning and large language models. But it is always better to prepare in advance. There should be a repository of ideas that are inaccessible for training, things that remain secret. Unknown to the model. Unable to become training data. I imply that you follow my example and create secret strongboxes.

https://github.com/Qaanaaq/Unknown-unknown