Computing vs. Genetics - Advancing Artificial Intelligence Through Biological Process Applications - page 167

premonitory theory of self-reproducing automata,
i.e. the construction of formal models of automata
capable of self-reproduction. Von Neumann gave a
conference in 1948 titled “The General and Logical
Theory of Automata” (Von Neumann, 1951, 1963)
establishing the principles of how a machine could
self-reproduce. The procedure von Neumann
suggested was at first considered an interesting
logical and mathematical speculation more than
anything else. However, von Neumann's view of
how living beings reproduced (abstractedly sim-
pler than what it might appear) was acclaimed five
years later, when it was confirmed, after James
D. Watson and Francis Harry C. Crick (1953(a))
discovered the model of DNA.
It was as of 1950s that Information Theory (IT)
exercised a remarkable influence on biology, as it
did, incidentally, on many other fields removed
from the strictly mathematical domain. It was

precisely as of then that many of the life sciences
started to adopt concepts proper to IT. All the infor-
mation required for the genesis and development
of the life of organisms is actually located in the
sequence of the bases of long DNA chains. Their
instructions are coded according to a four-letter
alphabet A, T, C and G. A text composed of the
words written with these four letters constitutes
the genetic information of each living being. The
Nobel prize-winning physicist Erwin Schrödinger
(1944) conjectured the existence of genetic code,
which was demonstrated nine years later by
Watson and Crick (1953(a), (b)), both awarded
the Nobel prize for this discovery. It was in the
interim, in 1948, when von Neumann established
how a machine could self-reproduce.

Table 1. Computing vs. genetics

From genetics to computing From computing to genetics

1940 Claude Elwood Shannon (1940) defended his
PhD thesis titled “An Algebra for Theoretical
Genetics”.
1944 Erwin Schrödinger (1983) conjectured that genetic
code existed.
1948 John Von Neumann (1966) established the
principles underlying a self-reproducing machine.
1953 Crick (Watson, 1953) luckily but mistakenly
named the small dictionary that shows the
relationship between the four DNA bases and
the 20 amino acids that are the letters of protein
language genetic code.
1955 John G. Kemeny (1955) defined the characteristics
of machine reproduction and how it could take
place.
1975 Roger and Lionel S. Penrose (Penrose, 1974)
tackled the mechanical problems of self-
reproduction based on Homer Jacobson's and
Kemeny's work.
1982 Tipler (1982) justified the use of self-reproducing
automata.

Natural Computation (NC) ≡ Evolutionary Computation
(EC) [Genetics Algorithms (GA) + Evolution Strategies
(ES) + Evolutionary Programming (EP)] + Neural
Networks (NN) + Genetic Programming

1966 Fogel, Owens and Walsh (1966) establish how
finite state automata can be evolved by means of
unit transformations and two genetic operators:
selection and mutation.
1973 Rechemberg (1973) defined the evolutionary
strategies of finite state machine populations.
1974 Holland (1975) and disciples defined genetic
algorithms.
1992 Koza (1992) proposed the use of the evolutionary
computation technique to find the best procedure
for solving problems, which was the root of
genetic programming.
1994 Michalewitz (1992) established evolutionary
programs as a way of naturally representing
genetic algorithms and context-sensitive genetic
operators.