
A NASA-funded artistic rendition of Neumann’s automaton that produce automaton, the original caption of which is: "proposed demonstration of simple robot self-replication", based on the Advanced Automation for Space Missions NASA/ASEE summer study held at the University of Santa Clara in Santa Clara, California, from June 23-August 29, 1980. [15]

A NASA-funded artistic rendition of Neumann’s automaton that produce automaton, the original caption of which is: "proposed demonstration of simple robot self-replication", based on the Advanced Automation for Space Missions NASA/ASEE summer study held at the University of Santa Clara in Santa Clara, California, from June 23-August 29, 1980. [15]

In science, Neumann automaton theory or “free energy automaton theory” is a thought experiment which argues that a robot or automaton, made of wires, electrical motors, batteries, etc., constructed in such a way that when floating on a lake stoked with component parts, it could reproduce itself (self-replicate); albeit one that could be made to work only if a source of free energy is available. [1] The theory is part thermodynamic-based, part electrical-chemical based, part computer engineering/computer programing based.

The self-replicative electrochemical automaton theory was first presented in 1948 by American chemical engineer, mathematician, and computer pioneer John Neumann.

Number of lectures
To note, as summarized by American mathematician and computer designer Arthur Burks (1966), Neumann seems to have given at least five or more different lectures (1948-1949), in which the various parts self-reproducing automaton theory are presented:

First: "Computing Machines in General"

Delivered: (add)

Second: "Rigorous Theories of Control and Information"

Delivered: (add)

Third: "Theory and Organization of Complicated Automata"

Delivered: (add)

Fourth: "The Role of High and of Extremely High Complication"

Delivered: (add)

Fifth Lecture: "Re-evaluation of the Problem of Complicated Automata: Problems of Hierarchy and Evolution"

Delivered: at the University of Illinois, December, 1949

Neumann then seems to have been in the process of assembling all of this into a book entitled The Theory of Automata: Construction, Reproduction, Homogeneity. Both the fifth lecture and the unfinished manuscript were published as the edited 1966 book by Arthur Burks.


A modern two-dimesional computer iteration version of Neumann's cellular automaton, according to which a binary signal is passed repeatedly around the blue wire loop, using excited and quiescent ordinary transmission states, during the course of which a confluent cell duplicates the signal onto a length of red wire consisting of special transmission states; the then signal passes down this wire and constructs or so-called self-replicates a new cell at the end (i.e. a new red arrow).

A modern two-dimesional computer iteration version of Neumann's cellular automaton, according to which a binary signal is passed repeatedly around the blue wire loop, using excited and quiescent ordinary transmission states, during the course of which a confluent cell duplicates the signal onto a length of red wire consisting of special transmission states; the then signal passes down this wire and constructs or so-called self-replicates a new cell at the end (i.e. a new red arrow).

Four talks
In 1953, Neumann delivered the Vanuxem Lectures at Princeton University, March 2-5, entitled “Machines and Organism”, which he intended to have published by Princeton University Press. Later, however, owing to, it has been argued, manifold activities and possibly failing health, he declined to publish these lectures. A summary of these lectures, however, can be found in a Scientific American article written by John Kemeny of Dartmouth. [13] A considerable portion of the first three talks, supposedly, are found in the 1958 posthumous book The Computer and the Brain. [14]

Health issues
In 1955, he was invited to deliver the Silliman Lectures at Yale University, during the spring of 1956. On 15 March 1955, Neumann was sworn in as one of the Atomic Energy Commissioner, and he and his wife moved to Georgetown in May. Three months later, in August, he developed severe pains in his left shoulder, and after surgery, bone cancer was diagnosed; in 1957 he reached his dereaction point (died). [12]

Thermodynamic information
In his third and fourth lectures, supposedly, Neumann outlined his ideas on "thermodynamic information", namely that, in his view "information is entropy" and that it is measured by the logarithm as follows:

$I_T = \log_2 N \,$

where N is number of alternatives, which tends to be two, i.e. a choice between a high or low signal pulse, in the simplified case, or binary digits (bits), involved in the measurement. He then calculated that the energy per elementary act of information (alternative decision and or transmission), which he equates to the following formula:

$E_I = kT \log_e N \,$

wherein k is the Boltzmann constant, T is the absolute temperature, and N is the number of alternatives. When N is two, i.e. the choice is between 0 and 1, i.e. low or high current, Neumann calculates what he calls the "thermodynamical minimum" which has a value of 3x10E-14 ergs at a temperature of about 300 degrees kevin, "per elementary act of information, that is, per elementary decision of a two-way alternative and per elementary transmittal of 1 unit of information." He then idealized the neuron to be a binary alternative reading vacuum tube and extrapolated the above logic to this scenario. [11]

Overview
The so-called “Neumann automaton theory” originated in a 20 September 1948 Hixton Symposium lecture, Pasadena, California, organized by American chemical engineer Linus Pauling, given by Hungarian-born American chemical engineer and mathematician John Neumann, during the course of which Neumann invented a famous thought experiment which illustrates the role which free energy plays in creating statistically unlikely configurations of matter. Neumann imagined a robot or automaton, made of wires, electrical motors, batteries, etc., constructed in such a way that when floating on a lake stoked with component parts, it will reproduce itself (self-replicate). Neumann, in his lecture, first compares computers to biological information processing systems then suggests a program to deal with “automata that produce automata”, the gist of which is captured in the following statement:

“Can one build an aggregate out of such elements in such a manner that if it is put into a reservoir, in which there float all these elements in large numbers, it will then begin to construct other aggregates, each of which will then at the end turn out to be another automaton exactly like the original ones?”

These so-called floating elements, as noted by American mathematician and computer designer Arthur Burks, who would later attempt to create a two-dimensional realization of Neumann's three-dimensional thought experiment premise, were spoken about prior to the lecture, by Neumann in the summer of 1948, as consisting of eight kinds of parts: a ‘stimulus organ’, a ‘coincidence organ’, an ‘inhibitory organ’, a ‘stimuli producer’, a ‘rigid member’, a ‘fusing organ’, a ‘cutting organ’, and a ‘muscle’. [10]

Free energy
The important point about Neumann’s automaton, as summarized by Lebanese-born Danish theoretical chemist and physicist John Avery, is that it would require a source of free energy in order to function: [2]

“The important point about von Neumann’s automaton theory is that it requires a source of free energy (i.e. a source of energy from which work can be obtained) in order to function. We can imagine that the free energy comes from electric batteries which the automaton finds in its environment. (These are analogous to the food eaten by animals.) Alternatively we can imagine that the automaton is equipped with photocells, so that that it can use sunlight as a source of free energy, but it is impossible to imagine the automaton reproducing itself without some energy source from which work can be obtained to drive its reproductive machinery. If it could be constructed, would von Neumann’s automaton be alive? Few people would say yes. But is such a self-reproducing automaton could be constructed, it would have some of the properties which we associate with living organisms.”

It does not seem to be the case, to note, that Neumann was thinking along the lines of “free energy”, as the term does not seem to be found in his lecture publications, but rather possibly this is but latter adumbrations and elaborations by other others, Avery being a possible case in point. Avery's notion of free energy as being equated with either a battery or, in the human/animal case, with "food eaten", however, is not the way in which "Gibbs free energy" is understood in chemical reaction theory, such as discussed in human chemical reaction theory.


A $1,450 dollar rare books copy of John Neumann’s high-sought 1966 The Theory of Self-Reproducing Automata. [8]

A $1,450 dollar rare books copy of John Neumann’s high-sought 1966 The Theory of Self-Reproducing Automata. [8]

In 1992, American computer scientist John Koza, in his Genetic Programming: On the Programming of Computers by Means of Natural Selection, for instance, stated that the rules of the Neumann milieu are such that there is “sufficient free energy and free matter available in the milieu to permit the creation of a copy of the desired thing.” This statement, however, does not seem to come from Neumann. This is exemplified by the fact that Koza, in his so-called "Entropy-Driven Evolution" chapter, culls his understanding of thermodynamics from the Shannon bandwagon, such as is exemplified by his statement: “entropy (i.e. information) drives the evolutionary process”. [6] Neither of these, information nor entropy, however, "drives" the evolutionary process, but rather the correct thermodynamic potential driving force, for the given reaction conditions of evolution, which are approximated as being isothermal and isobaric and freely going, is Gibbs free energy. [7]

Supposedly, to note, Neumann, in his first design attempt, disregarded the fuel and energy problem, planning to consider it later, perhaps by introducing a battery as an additional elementary component. It is argued that the full kinematic machine was never realized in hardware, not even fully designed, due in part to Neumann’s premature reaction end (death) in 1957. [3]

Publication
In 1951, Neumann's original 20 September 1948 lecture was expanded or rather elaborated on in more detail, as Neumann says, into the form of a 39-page chapter publication "The General and Logical Theory of Automata". Neumann, supposedly, did not but complete and publish any further work on this topic.

In 1966, American mathematician and computer designer Arthur Burks published Theory of Self-Replicated Automata, an edited, what seems to be, collected works set on the bulk of Neumann's automaton theories. [5]


An artistic rendition of the 1980 conceived Neumann automaton theory based self-replicating lunar space station, a premise about which NASA spent 11.7 million dollars investigating. [9]

An artistic rendition of the 1980 conceived Neumann automaton theory based self-replicating lunar space station, a premise about which NASA spent 11.7 million dollars investigating. [9]

NASA
In the summer of 1980, at the request of American president Jimmy Carter, 11.7-million dollars were spent on a study to investigate a potential realization of Neumann’s self-replicating automaton theory for a possible lunar factory system, which would theoretically be capable of exponentially increasing productive capacity and, in the long run, exploration of the entire galaxy within a reasonable timeframe. [9]

References
1. (a) Neumann, John von. (1963). "Probabilistic Logic and the Synthesis of Reliable Organisms from Unreliable Components", in Collected Works (A. Taub editor), Vol. 5, pgs. 341-47. MacMillian, New York.
(b) Neumann, John von. (1966). Theory of Self-Replicating Automata (scanned) (editor: Arthur W. Burks). University of Illinois Press.
(c) Avery, John. (2003). Information Theory and Evolution (automaton, pg. 89 and ch. 8). London: World Scientific.
2. Avery, John. (2003). Information Theory and Evolution (automaton, pg. 89 and ch. 8). London: World Scientific.
3. Freitas, Robert and Merkle, Ralph C. (2004). Kinematic Self-Replicating Machines (pg. 8). Landes Bioscience/Eurekah.com.
4. Wiener, Norbert and Schade, J.P. (1965). Cybernetics of the Nervous System (pg. 29). Elsevier Pub. Co.
5. (a) Neumann, John von. (1966). Theory of Self-Replicating Automata (scanned) (editor: Arthur W. Burks). University of Illinois Press.
(b) Arthur Burks – Wikipedia.
6. (a) Koza, John R. (1992). Genetic Programming, Volume 1, On the Programming of Computers by Means of Natural Selection (§14: Entropy-Driven Evolution, pgs. 395-418; Neumann, pgs. 648-49). MIT Press.
(b) John Koza – Wikipedia.
7. (a) Thims, Libb. (2012). “Thermodynamics ≠ Information Theory: Science’s Greatest Sokal Affair” (url), Journal of Human Thermodynamics, 8(1): 1-120, Dec 19.
(b) Thims, Libb. (2007). Human Chemistry (Volume One). Morrisville, NC: LuLu.
(c) Thims, Libb. (2007). Human Chemistry (Volume Two). Morrisville, NC: LuLu.
8. Neumann, John. (1966). The Theory of Self-Reproducing Automata (abs). Bauman Rare Books.
9. Freitas, Robert A. and Gilbreath, William P. (1982). Advanced Automation for Space Missions: proceedings of the 1980 NASA/ASEE summer study sponsored by the National Aeronautics and Space Administration and the American Society for Engineering Education held at the University of Santa Clara, Santa Clara, California June 23-August 29, 1980 (url). NASA, Scientific and Technical Information Branch.
10. Mirzoeff, Nicholas. (2006). “Network Subjects: or, The Ghost in the Machine”, in: New Media, Old Media: a History and Theory Reader (§23, pgs. 335-46) (editors: Wendy Chun and Thomas Keenan) (pg. 237). Taylor & Francis.
11. Neumann, John. (1949). “The Role of High and of Extremely High Complication”, Fourth Lecture, in: Theory of Self-Reproducing Automata (editor: Arthur W. Burks) (pgs. 64-87). University of Illinois Press, 1966.
12. Goldstine, Herman H. (2001). The Computer from Pascal to von Neumann (pg. 277). Princeton University Press.
13. Kemeny, John. (1955). “Man Viewed as Machine”, Scientific American, 192:58-67.
14. (a) Neumann, John. (1958). The Computer and the Brain. New Haven; Yale University Press, 2012.
(b) Goldstine, Herman H. (2001). The Computer from Pascal to von Neumann (pg. 277). Princeton University Press.
15. Artist. (1982). “Image: Neumann self-reproducing automaton”, NASA Conference Publication 2255 (1982), based on the Advanced Automation for Space Missions NASA/ASEE summer study Held at the University of Santa Clara in Santa Clara, California, from June 23-August 29, 1980.

Further reading
● Neumann, John. (1951). “The General and Logical Theory of Automata”, Cerebral Mechanisms in Behavior; in: Modern Systems Research for the Behavioral Scientist (editor: Walter Buckley) (§21, pgs. 97-108). Aldine Publishing Co., 1969; in: “The General and Logical Theory of Automata”, in: Cerebral Mechanisms of Behavior: the Hixon Symposium (pgs. 1-#). California Institute of Technology.
● Penrose, L.S. (1959). “Self-Reproducing Machines” (pdf), Scientific American, 200(June):105-14.
● Cooper, Necia G. (1983). “From Turing and von Neumann to the Present”, Los Alamos Science, Fall.

External links
● Von Neumann cellular automaton – Wikipedia.