My assumption is that a machine learning network is capable of taking advantage every input thrown at it.
So with a universe of networks I wonder. Could this mean that the universe takes advantage of things not so obvious. Like the in between space objects the atoms create in a material.
So if atoms have properties. Could empty space objects with a somewhat defined border also have properties?
Are there more not so thought of objects?
It could be that the object have important information properties. Like the border of an object is a object you can take information from. That is. Statistics about the border. Something similiar you do in machine learning where you might take the avarage smoothed border curve as input.