Editorial: Whole Brain Emulation seeks to Implement a Mind and its General Intelligence through System Identification

Abstract

Whole brain emulation (WBE) is a systematic approach to large-scale neuroprostheses with theintent to replicate the functions of a specific mind in some other operating substrate. The engineeringpractice of system identification can be applied in a way that makes this big problem a feasiblecollection of connected smaller system identification problems to solve.Whole brain emulation is an essential goal for neuroscience. Following Richard Feynman’sfamous 1988 Caltech chalkboard quote: “What I cannot create, I do not understand.” To create orbuild a human mind we need models, a combination of building blocks with processes. When weexplain something that is observed, e.g., mental functions and behaviors, we strive to make thatpredictable within constraints that satisfy our interests: We create boundaries, we measure withinthose well-defined outlines, and then we use those measurements to derive model processes enablingoutcome prediction. Within the defined system outlines of our model, taking into account definedsets of signals, we mathematically describe interactions (which may be expressed in informationtheoretic terms).Every aspect of modern science relies on creating representations of things. In each case, wefocus on the signals and the observables (or behavior) that interest us. Then, we try to interpret interms of functions what the system processes are doing. Where brain functions are concerned, somecognitive prosthetic work, such as the pioneering efforts of the labs of Theodore W. Berger at theUniversity of Southern California, has managed to carry out these steps and produced successfulexperimental results (Berger et al., 2012). Berger’s team has developed and tested an experimentalhippocampal neural prosthetic that is implemented on a bio-mimetic chip. A transfer functionwas identified and used to replicate the operational properties of biological neural circuitry in aregion of the rat hippocampus known as CA3. In experiments, the prosthesis is able to reproducethe way in which input to the region is turned into output from that region. This method ofdeveloping neuroprostheses, with demonstrated success in rats, is presently being tested in primates(Marmarelis et al., 2013).