Artificial intelligence and neural networks - The Computing Universe: A Journey through a Revolution - page 274

the cerebral cortex , a thick layer of neural tissue that covers most of the brain (see
Fig. 13.13 ). The name cortex comes from the Latin word for the bark of a tree, but
in this case it means the outer layer of an organ. The cortex is deeply folded and
ribbed because such folding maximizes the amount of brain surface that can fit
into the limited space of the skull. More than two-thirds of the surface area of a
human brain is buried in these folds, called sulci . The cerebral cortex plays a key
role in memory, perception, thought, language, and consciousness.
The nineteenth century brought rapid progress in biological science thanks
to the wide use of microscopes. Theodor Schwann and Matthias Schleiden had
suggested cell theory, according to which all living organisms are made up of
cells, in 1838. But not all scientists were convinced that cell theory applied to
brain tissue. As a result, many scientists experimented with different chemi-
cal substances for coloring the brain tissue so that individual cells would be
made visible. A Spanish physician, Santiago Ramón y Cajal improved on a cell-
staining method originally developed by the Italian doctor Camillo Golgi, and
used this new technique to investigate the central nervous system of many
living creatures. It was Ramón y Cajal's work that first revealed the complexity
of biological neural networks. He wrote:

Fig. 13.13. A diagram of the cerebral
lobes of a human brain: frontal lobes
in pink, parietal lobe in green, and the
occipital lobe in blue.

What beauty is shown in the preparations obtained by the precipitation of
silver dichromate deposited exclusively onto the nervous elements! But, on
the other hand, what dense forests are revealed, in which it is difficult to
discover the terminal endings of its intricate branching.… Given that the
adult jungle is impenetrable and indefinable, why not study the young forest,
as we would say in its nursery stage. 18

We now know that neurons consist of a cell body or soma with two types of
nerve fiber growing from the cell, dendrites and axons . The cell body contains
the genetic information and the molecular machinery required for the func-
tioning of the neuron. The role of the dendrites is to receive electrical or chem-
ical signals from other neurons and provide the input to the cell of the neuron.
The axon, usually much longer than the dendrites, carries nerve impulses from
the cell body to other neurons. Ramón y Cajal also suggested that these signals
always flow in one direction, from the dendrites of the cell to the axon, and that
the axon is connected to dendrites of other cells by structures called synapses
(see Fig. 13.14 ). The word synapse comes from the Greek words syn , meaning
together , and haptein , meaning to clasp . Golgi and Ramón y Cajal were awarded
the 1906 Nobel Prize in physiology or medicine “in recognition of their work
on the structure of the nervous system.” 19
The number of neurons in the brain varies widely from species to species.
The human brain is believed to contain more than eighty-five billion neurons,
while the brain of a cat has only one billion and a chimpanzee about seven bil-
lion neurons. In addition to these vast numbers of neurons, the brain has an
even larger number of synapses. Each human neuron has, on average, seven
thousand synaptic connections to other neurons. There are many different
types of neurons, and we will describe only how a “typical” neuron functions.
The incoming signals reaching a neuron from all of its dendrites are collected
and processed inside the cell body. Any output signal resulting from this input

B.13.8. Warren McCulloch (1898-
1969) was an early pioneer of AI.
With Walter Pitts he proposed the
first mathematical model of a neural
network. John von Neumann was
very impressed by the McCulloch-
Pitts model and the paper and its
physics terminology influenced him
as he wrote the EDVAC draft report.