Great question Hixy, and great & long) to answer.
We did it hierarchically. eg
Top down.
Does intelligence exist anywhere? How does it work? If you cant know how it works, what are your best guesses?
Bottom up
What can you build?
eg a neuron? a network?
We thought a general working definition of intelligence was useful, so defined it as general problem solving.
Also a unit of intelligence was thought useful too, so we called it a joule ie
1joule of intelligence - and defined it as the movement of one digit toward the goal of problem solving.
A goal is a finishing point...sometimes there be no end to it...you keep going, but it is still a goal)
Next we had to look at how we would construct it and our first attempt was using computers ie Intelligence -Memory modification @ speed (I=MS) This satisfied solving 'learning'.
Unfortunately you would run into complexity and memory storage limits like this so we had to redesign.
Deep Learning doesn't reply on memory but solves each task anew which is fun. It uses back propagation, and the human brain doesn't so the principles diverge.
Hard to see how far neuromorphics will go, except that intelligence is not only a product of the system, but of the environment which modifies it this way.
It's probably great to think like Pythagorus and Plato that everything is describable in number, has order, and is capable of measurement.
If you're going for a build, remember you have to be able to turn it off/make it safe,
and it's too easy to get unnecessarily delayed in philosophy.