Abstract
Alcibiades Malapi-Nelson has written a dense historical, social, and philosophical or intellectual study of the short life cycle of what he calls the “metaphysical research programme” of cybernetics. I will start off with a brief, superficial, and simplified overview of Malapi-Nelson’s quasi-sociological and quasi-intellectual history (or, “rational reconstruction,” a term he does not use but that is often paired with the term, “metaphysical research programme”) of the life, death, and the apparent recent afterlife (or perhaps, “resurrection”) of the sputtering intellectual flame of cybernetics as both the science of control and the theory of self-governing, feedback mechanisms. From the time of the early World War II (WWII), war years both in the United States and in the United Kingdom—sometimes independently and sometimes conjointly—until the early post war years, from 1942 to 1952-1953 roughly—cybernetics under the immediate but indirect influence of Turing’s virtual machine, and with the push of Norbert Wiener; with the counterbalance of the versatile brilliance of John von Neumann, sometimes from the side and sometimes from front and center, in the United States; and with the driving force of W. Ross Ashby’s insistent and persistent mind, cybernetics was born, developed, and ultimately faded away due mainly to an internal instability. It tottered, wobbled, and spun down and ultimately “imploded” over its main principle of developing material models of theories, including its own. However, near the end of its life, just as it began to fade away into the intellectual ether of historically interesting but self-refuting theories, the advocates of cybernetics began to loosen the materiality requirement and began to replace that requirement with the virtuality requirement: can all machines be simulated? For instance, can all machines be simulated within a computer simulator? Or even more abstractly put: can every theoretical model of machines be virtually simulated within a theoretical model of computation such as the Universal Turing Machine? Such questions allowed cybernetics to find an afterlife, and more so, in the current nascent life of Transhumanism. Transhumanism refers to the new ideology forming around the belief in the potential replacement of our species by a new form of humanity that is supposedly evolving from the current technological developments in microbiology, and various other budding technologies, that include enhanced prosthetic devices, artificial general intelligence (AGI), nanotechnology, and cryonics.
There is nothing too intellectually hair-raising up until the last sentence of my condensed version of Malapi-Nelson’s very thorough historical, intellectual, and sociological study. One may wonder: what has Transhumanism to do with an intellectual and social history of cybernetics? Rather than discuss how current nanotechnology, bioengineering, recombinant DNA projects, and so forth simply build on the legacy of cybernetics in terms of treating everything as mechanisms, even, micromechanisms that include feedback, self-governance, and self-reproduction—as autonomic mechanisms; Malapi-Nelson treats these new fields of research as part of an intellectual tsunami wave generated by the search for Transhumanism as the drive to defeat mortality and create a new superspecies that replaces humans as we have been and are now. According to Transhumanism, due to our own design and intervention, humans will evolve or morph into a superspecies. The new version of humanity will surpass current humanity. The new version of humanity will achieve superintelligence: perhaps in the form of the so-called, singularity of computer systems, or perhaps in the form of human-machine symbiosis (of cyborgs, not mentioned in this book); or perhaps in the form of designer humans with intentionally designed genes that provide self-healing mechanisms, superintelligent brains, and superstrong musculature. According to Transhumanism, we will create a designer version of the species, Sapiens: version 2.0.
From the side, and perhaps, from the outside of this new avant-garde, with Malapi-Nelson’s form of intellectually advancing the intellectual advancers, one might pose some more moderate version of the renaissance of cybernetics. Cybernetic notions infuse the new developments in technology, biology, and even mathematics as a branch of computer science and computer modeling. Also in theoretical quantum physics, cybernetic notions infuse the information-theoretic model for Quantum Mechanics and Quantum Computing, and infused from the start Hugh Everett’s multiverse view of Quantum Mechanics where both the observed and observer are interacting cybernetic mechanisms. Specifically, as with some of the pioneers of the personal computer who were influenced by cybernetics, not discussed in this book—such as J. C. R. Licklider (1915-1990) in the 1960s and later, in the 1970s, those who worked in Xerox Palo Alto, on the development of “user-friendly” ergonomics or interfaces, were also influenced by cybernetics. They developed personal computer technologies such as the mouse, haptic interfaces and technologies such as the touch screen, and graphic iconic representations of computer functions, as part of the deployment of the family of cybernetic concepts related to the augmentation of human intelligence, and augmentation of human sensory or other physical functions. Marshall McLuhan, not mentioned in this book, early on tacitly talked cybernetics when he talked about how machines can become extensions of our hands, arms, eyes, ears, and even cognitive functions. The extreme version of McLuhan’s theory of machines as extensions and transformations of human abilities verges on Transhumanism: the electronic media not only create a new ecology that encourages the use of nonlinear, multi plural, multitasking cognitive functions, but also modify our brains to adapt to the new electronic ecology of virtuality—not only symbols, but also simulations including holographic simulations.
The short of it is the question of Transhumanism as opposed to augmentation: do new technologies, whether cybernetic or not, and even if wholly cybernetic, push us in the direction of a posthumanist world of new super creatures that replaces the species of Sapiens? Or, is our technological reality a bit more humdrum? We are not sure what direction our technological, biological, and other innovations—social, institutional, political—will take. Will we end up merely, at the best, improving our physical and mental skills and abilities? Or at the worst, will we end up creating only superbugs where the cure for the disease is the extinction of the species, Sapiens?
What I have done so far is set the stage for the controversy inherent in this book: For instance, those on the outside of the new school of Transhumanism, can validly argue that the argument for the transcendence of humanity (Transhumanism) is invalid, because the argument for such a transcendence actually is not Darwinian but Lamarkian, and so based on a false premise: a change in technologies and sciences—in our ecology—leads to the inheritance of the new adaptations to this environment; in other words, a top-down imposed inheritance as opposed to the Darwinian bottom-up trial and error process of the elimination of chance or otherwise changes in our genetic material. Even if and when we are successful in modifying our genetic material, such as already happened with recombinant DNA experimentation on various bacteria, we may end up with monsters or duds, as opposed to a superspecies, such as many of the failed results with recombinant DNA experimentation on bacteria that are supposed to eat oil spills. One might find distracting the entire discussion of Transhumanism in this book, confined to the last few chapters starting from most of chapter 9 (227ff.), all of chapter 10, and the last punch up paragraph of the book’s Concluding Remarks, chapter 11. I want to leave open the question Malapi-Nelson’s discussion of Transhumanism raises, that some will see as out of alignment with the main focus of the book on the life cycle(s) of cybernetics: can we predict even if we intend our innovations in science, technology, social institutions, and so forth to create a new and improved version of humanity, that we will achieve what we expect? Or will we produce unintended and unwanted effects such as we did with superbugs; the pollution of our soil, waters, and air such that those populations in Brazil, Russia, India, and China (BRIC) economies suffer unhealthy and shortened life spans; the endangerment and extinction of many species; and the increasing numbers of under- and unemployment in the various rust belts of our global and technological, so-dubbed, Information economies?
Now that I have identified and I hope closed into a box, what I think could be a distraction for the reader, I want actually to focus on the main thrust of the book: the life cycle of cybernetics as thoroughly explored in the first eight chapters. This is not to say: do not read the rest of the book. I just provide an advisory for the book, to those readers who are not impressed by the science-technology fantasies of comic novels about superheroes and super demons, and their afterlife in animated movies, and multiplayer online games. Those readers too, might not be impressed by the science-technology fantasies of Transhumanism, even if Transhumanism has a legacy that begins with Duns Scotus, Francis Bacon, and Giambattista Vico (237ff.). Isaac Luria, the instigator of Lurianic Kabalism, can also be included in the prehistory of Transhumanism though Malapi-Nelson does not mention him. According to Lurianic Kabalism, we have an imperfect world including imperfect humans, that requires human intervention and perfection, especially of humans. Lurianic Kabalism apparently influenced Bacon to develop a version of scientism where science and technology can be developed to achieve the ultimate perfection of humanity and human society. But a distinguished prehistory for an intellectual movement does not legitimize, nor intellectually dignify the movement. Rather, the Transhumanist movement may actually turn out to be an unintended and unwanted intellectual and political offspring, of those in our intellectual history who advocated for the perfection and completion of our imperfect world and humanity.
How does the life cycle of cybernetics begin? It begins with a very serious and well-known problem with mechanisms that depend on feedback, such as anti-aircraft guns, and audio amplification systems. The screeching and wailing sound emitted by sound systems where unwanted feedback occurs is hardly often deadly, and most often only annoying. However, how do we control feedback so that we can filter in only wanted feedback that for instance, aims the anti-aircraft gun at the moving airplane, and allows the gun to shoot down the enemy airplane rather than wobble unsteadily and misfire at the enemy plane? This was the problem that initially moved Norbert Wiener to come up with a specific and then later general solution to using feedback for accurate machine performance that he later dubbed Cybernetics. Cybernetics ultimately evolved through a trial and error method with model mechanisms that use feedback, to produce working machines including weapons, such as the submarine torpedoes that use echoes and the Doppler effect, to guide the weapon toward the target (26).
Apart from the intellectual and practical background (better auto-guided artillery for use in WWII) explored by Malapi-Nelson, Malapi-Nelson also explores the deep intellectual background including the deep mathematical background that stems from Cantor, Frege, Hilbert, Gödel, Church and Turing—in other words mathematical formalism—in chapter 4 (81ff.). But leading up to this investigation is Malapi-Nelson’s exploration of the standard explanations of the failure of cybernetics. Cybernetics had a great boost in the relatively informal intellectual associations that sprang up around those in the United States and in the United Kingdom who were beginning to research the issues. The key cybernetic researchers in the United States used the “Macy Conferences” (33ff.), for their intellectual association and the key researchers in the United Kingdom used the “Ratio Club” (56ff.). However, the sine curve of birth, growth, and decline was steep. It only took 12 years for both associations to dissolve. Why? Why the short life span when hopes were great, and the associations included extremely brilliant and versatile polymathic thinkers? This is Malapi-Nelson’s central question. Not: why did cybernetics attract so many smart people? Not: how did cybernetics solve, if it did, major problems in computing, artificial life, and artificial intelligence (AI)? Not: how did cybernetics become a model for technological innovations in ergonomics, machine learning, and the use of technology, let alone the mathematical modeling of control and automated systems? Rather: why did cybernetics crash or at the least, fade away as did many other bandwagon intellectual phenomena?
Malapi-Nelson explores in the subsection “Traditional Explanations for the Collapse of Cybernetics” (67ff.) why such things as McCarthyism regarding the suspicious behavior of Wiener, in the eyes of the communist witch hunters, did not gain needed funding, or the eclipse of Wiener’s program by leading AI researchers who adopted the symbolic, heuristic, model for developing AI as opposed to the more (analogical, I think) feedback mechanism, and the more polycentric bottom-up model approach of simulating neural networks in brains, of Wiener and his colleagues, especially Ashby in the United Kingdom. Not even a falling out between von Neumann and Wiener, nor the machinations of Wiener’s manipulative spouse adequately explain the collapse of cybernetics. Basically, though each of the above explanations is sufficient unto itself, none is necessary—none gets at the heart of the matter. The heart of the matter is the basic question facing every intellectual program that attempts to maintain intellectual honesty: has the program been able to adequately, if not completely, solve its own intellectual problems according to its own agenda of problems and intellectual requirements? In other words, has it been able to achieve, in cybernetic terminology, “homeostasis”? Or were there unresolved and unstable elements within the program of cybernetics itself that caused it to falter and tumble to a crash?
For the reader who is eager to get right to the answer without exploring in detail the intellectual background, immediate problems, requirements and agenda, I suggest skipping to chapter 8, where the answer to the internal intellectual problem of how cybernetics spun out of control and flew away into intellectual oblivion is given, while allowing that it was deserving of rebirth in Transhumanism (as discussed above and not further to be discussed in this review). In the rest, I will come to the bottom line of why and how Malapi-Nelson argues for the internal dissolution of cybernetics—as if apoptosis occurred in the social and intellectual cybernetic organism.
The last sentence of chapter 7 actually concisely sums up the internal reason for the collapse of cybernetics: “Once physical instantiation is out of the equation, cybernetics is left with little that would make it different and unique, let alone revolutionary” (201). The “materiality” of machines, and the search for mechanisms as the mode of cybernetic explanations, and so the representation of explanatory mechanisms in material form, was the raison d’être for cybernetics: both its goal as a worldview (“metaphysical research programme”) and its requirement for a scientific or empirical endeavor as a means of testing theories or theoretical explanatory systems. If one can build the theory, and the constructed mechanism produces the results that one is attempting to explain or predict, then the theory has passed its test. This requirement for a physical or material model, in current systems development, is called “proof of concept.” If the prototype computer or even business/social operation fails, then the concept is either modified or abandoned. Furthermore, when the mechanism does not produce the predicted and expected result, but another unwanted result, then the theory fails; but either one tries again with a new and improved mechanism, or one attempts to figure out what actually the theory is explaining and the mechanism is doing. This might turn out to be a serendipitous discovery, such as the mechanism that produced penicillin, or for computer systems developers, a system containing hidden features until they are unexpectedly found by hitting a strange sequence of keys or inputting data in the wrong field or in the wrong format, producing an interesting result that then is called an undocumented feature and propagated by word of mouth and social media.
When one withdraws the goal and the requirement for materiality, for constructing actual mechanisms as a representation of abstract explanatory theories, one pulls out the cybernetic rug from under one’s feet. Here then is the key question of the book, given that Malapi-Nelson is excluding externalist explanations for the collapse of cybernetics, and only seeks internal mechanisms: to what internal intellectual struggle did the intellectual leaders of cybernetics respond by withdrawing the materialist metaphysical research programme that was to them the fundamental principle of cybernetics—the very core of the matter of cybernetics that if taken away would leave the intellectual leaders of cybernetics floating without direction and without control in intellectual space? The intellectual struggle actually is very simple: both Ashby and von Neumann found independently that building physical models of their respective attempts at developing a mechanism that explains reproduction, to understate the problem, is difficult. Ashby wanted to develop a mechanism for demonstrating adaptability and learning:
. . . the chaotic behavior of the newer more complex automaton never seemed to reach, let alone settle . . . despite 2 years of careful automation. Ashby, discouraged, abandoned the project and focused on the theoretical exploration of “cybernetics of cybernetics” . . . engaging in projects without physical grounding and effectively going under the radar . . . until the end of his life. (212)
Similarly, von Neumann’s failure to develop a physical model of his theory of self-reproducing automata, ground the project to a halt:
Issues of solderability, separation, energy sources, friction, heat material fatigue, and so on crippled the model right from the start . . . this impasse pushed von Neumann to consider an alleged equivalent of this self-reproducing entity, but now living in logical space. . . . After jumping out of the physical grounding into the abstract realm . . . the project failed to garner attention. Shortly after, von Neumann died. (216)
One could say, cybernetics moved from the materialist outlook to an idealist outlook: when the intellectual leaders of cybernetics found physical reality intractable to their research program, they transformed their research program from an empirical science into a branch of mathematics and logic.
However, at this point, I think it would be fair to point out that currently, AI has abandoned the symbolic approach that AI researchers developed to replace the cybernetic models based on neural functioning, and has returned to neural functioning as the main model for AI development: multimodal, parallel processing, with bottom-up learning through making and remaking neural-like connections. Whether we call it “cybernetics” or “deep learning,” as the new term of fashion, neural modeling is based on the cybernetics work of McCullough and Pitts (see especially 31ff., 68ff., 186ff., 194ff.) that was ultimately rejected by Marvin Minsky and Seymour Papert (outsiders to cybernetics) and the ultimate insider, John von Neumann.
Moreover, though not mentioned by Malapi-Nelson, the neural-like network models for AI have been recalled and redeveloped by current researchers in AI. Cybernetics has an afterlife in the materiality of the currently developing deep-learning or machine learning in AI and Artificial General Intelligence (AGI) mechanisms—computer simulated game playing (with other computers) and prototype mobile robots and robotic arms for testing neural networks that auto-learn both specialized and generalized intelligent functions or activities.
Malapi-Nelson could have ended the story here, and concluded with a lessons learned chapter. Rather, at this point, as discussed above, Malapi-Nelson brings into play the new form of Utopian thinking called Transhumanism. The rest of the book looks at how various new biotechnologies and new applications of microbiology can provide a renaissance of cybernetics. The question then becomes for Malapi-Nelson, how can we rebuild the mechanisms governing human development to move in the direction of achieving immortality, superintelligence, and super abilities? But one could raise a more modest question, how can we improve the current mechanisms we use, both institutional and physical, so that we can mitigate the unexpected damage done to our planet and to fellow humans?