Here I talk about implementation details involving the conceptual approach for the intelligence model laid out in the previous post.
There are two different kinds of inputs to the intelligence during each time step.
Following standards in neural networks, I will model sensory inputs with a vector of positive floating point numbers, and the cost is also a floating point number.
However, to make this less reliant on outside parameters, and therefore easier for a programmer to work with, there is no particular limit to the magnitude of the cost or senses, as long as they make sense in relation to each other and over time.
The outputs are actions, motions, etc, that effect the “outside” world. This is also a vector of floats, but absolute magnitude is going to be relevant here. The interpretation is “the extent to which something moves.” So when a programmer is making a robot, then