Synthetic neural modeling applied to a real-world artifact - by Gerald Edelman et.al (1992)

"The Neurosciences Institute was a unique place. The director was Nobel Laureate Gerald Edelman. In addition to his work in immunology, which led to the Nobel Prize, he introduced a theory of the nervous system called Neural Darwinism: The Theory of Neuronal Group Selection (Edelman, 1987, 1993). The theory suggested that there was selection of neural circuits during development through synaptic pruning, and selection of groups of neurons during adulthood through reentrant connections. Important for neurorobotics was the notion of value systems to tie environmental signals to neuronal groups, which led to the selection of behaviors important for survival. Because of this linkage, or as Edelman would say, “The brain is embodied, and the body is embedded in the environment,” their group developed the Darwin series of Brain-Based Devices (Reeke et al., 1990; Edelman et al., 1992). Another phrase that drove this work, was “the world is an unlabeled place,” which meant that perceptual categories must be selected through experience, rather than supervision. These Brain-Based Devices were robots3 with large-scale neural networks controlling their behavior (Figure 2). However, these were not the feedforward input layer→hidden layers→output layer neural networks that were popular then and became the deep neural networks of today. The Brain-Based Device’s neural networks had anatomical details that resembled biological neural networks. There were sensory streams, top-down connections, long-range connections between regions that were bi-directional, as well as local lateral excitation and inhibition within brain regions."
from https://www.frontiersin.org/articles/10.3389/fnbot.2018.00042/full

#NeuroScience #ALife