Sher

We really know of only a single intelligence abstraction approach that truly does work, the one based on the interconnection of spatio-temporal signal integrators in a vast graph: Neural Network. We also know of only one method that was able to generate such abstracted intelligence: Evolution. The proof that this abstraction and this generative method works is us, you and I, the result of billions of years of trial and error. There is nothing mystical about the human brain, it is but a vast graph of signal integrators, carved out in flesh through billions of years of evolution. In this paper we discuss: intelligence abstraction based on neural networks, complex-valued artificial neurons and their computational potential to be equivalent to biological ones, the approaches that could result in the generation of such intelligent graphs of interconnected complex-valued neurons, an architecture of infomorphs whose brains are complex-valued neural substrates, and why an ALife approach on high enough granularity level is our best chance of evolving organisms that are truly intelligent.