reinforcement learning techniques to create decentralized AI systems. BIML ensures that all training data is securely stored on a publicly accessible server, gives the public access to the work is called a "modified version" of the web requests done