Chapter
1
My Thesis
Introduction. This thesis explores how the natural progression in the development of information processing systems [1-16] will come to comprise millions of interconnected nanoelectronic components, [17-33] and how the ever expanding needs of future information systems will drive the evolution of nanodevice structures into a state of complexity sufficient for emergent behavior. [34-55] As these information processing systems become more involved in autonomous functions, they will be required to interface more directly with their local environments and to make decisions and change their behavior based on their received inputs. A neural processor, also referred to as a neural computer or neural network, is the most favorable processing system with this potential. [56-73] Nano-microprocessors, or “nanoprocessors,” will function as artificial neurons for these systems. Clusters of nanosystems consisting of these new artificial neurons and their corresponding nanoelectronic devices will enable evolution of the conditions required to produce a state of complexity both necessary and sufficient to lead to emergent behavior. In addition, the ever-minute dimensions of nanodevices produced by the self-assembly of molecular structures from individual atoms [74-93] will dwell within a realm dominated by quantum effects. Properly designed, the patterns of these nanostructures can be made to interact in a quantum entanglement to produce self-organizing quantum coherence states [94-99] on the order of tens of milliseconds. The collapse of this coherence state will likely produce a moment of "awareness" within the processor. [100-110] An ordered stream of quantum coherence collapses is speculated to produce consciousness. [111] The macrosystems resulting from the profuse clusters of these complex but now conscious nanosystems will ultimately lead to an emerging sentience – a product of consciousness and a precursor to true intelligence. [112-123] This dissertation will develop the philosophy of the likely necessary conditions for sentience to emerge from the ever-increasing complexity evolving from the continuing research into nanotechnology, specifically in the field of nanoelectronics.
The hypothesis of this dissertation is that the continuing thrust toward ever greater miniaturization of electronic microprocessors into the realm of nanoprocessors, coupled with the development of more capable “training” of neural processors, will enable computer capability that will evolve into a state of complexity possessing the conditions necessary for artificial sentience.[1] The dissertation builds upon the recent efforts in the fields of complexity and nanoelectronics to develop a theory for the necessary conditions that must be present in a complex artificial system in order to induce sentience. The original contribution is the development of this theory, and its presentation is in the form of a philosophy[2] based upon information research that describes the probable scenario of how such a state of complexity will come about, much through the overt actions of computer scientists responding to the economic drivers [124-130] calling for more sophisticated applications and more intelligent systems.
The
Venn diagram below (Figure 1.1) describes pictorially the interaction between
nanoelectronics and complexity to produce the necessary conditions for
sentience. As nanotechnology development
continues to progress and as more of this effort is directed toward
nanoprocessors and neural networks for information processing, by physical
necessity the computing systems will become more complex. As more capable devices are developed to
merge and integrate with these new systems and as greater system autonomy in
response to environmental changes is implemented, complexity will not only
naturally evolve for a myriad of reasons but it will also become a major
issue. But lying beneath this state of
complexity and highly sophisticated nanotechnology will be certain asymmetric
far-from-equilibrium physical conditions, which if present with the correct
balance, will be the required catalyst to bring about artificial sentience – a
state of awareness necessary for adaptive, brain-like intelligence. Because chemical synthesis is likely to be
the means for producing the required nanoelectronics and their
interconnections, a more appropriate descriptor for this emerging phenomenon is
"synthetic sentience." That
term will be used in this thesis.
Unlike artificial intelligence, wherein a computer is programmed via software and algorithms with characteristics that imitate human intelligence, [131-138] synthetic sentience becomes that intelligence. Artificial intelligence is accomplished through computer-processing functions that accumulate and manipulate data represented as symbols – a binary, digital operation. In addition, the computers within which these symbols are manipulated are closed systems where data transfers from its memory to its central processing unit and returns. But as we will see in the later chapters, the human brain is not a computer. It is neither binary nor digital, nor is it a serial processor. The computer analogy of the mammalian brain can be taken only so far. If a man-made information processor is to acquire brain-like characteristics, then how it approximates brain functions must be approached from a different paradigm. Furthermore, computers of this sophistication are far beyond the machine stage. They are a much more complex entity, and referring to them as “machines” is fundamentally inaccurate.
Complexity. But the evolution of nanotechnology, or more specifically, nanoelectronics is only part of the story, and its relationship to an emerging sentience can not really be understood in the absence of complexity. [139-141] And “complexity” when used in this connotation must not be confused with “complicated.” It is much more than that. With complexity comes the far-from-equilibrium "edge of chaos" that is conducive to the system self-organization necessary to produce higher-order structures. [142-150]
The whole of Nature itself is greater than the sum of its parts. [151-155] And it can not be understood in total by dissecting it into its individual components; i.e., via reductionism. Only parts of Nature can be learned through the classic reductionist approach. [156-159] The paradigm of reductionism must not be underrated, however, as it alone has been the most predominant contributor to our present-day understanding of Nature. No other methodology has so successfully pulled our knowledge of the physical world out of the Dark Ages of ignorance. But as one stands back and looks at natural systems as a whole, one finds that attempts to understand these systems from a purely reductionist approach leads to problems. A "simple" DNA molecule, for instance, can be disassembled into its four primary chemical compounds and broken further into the atoms that make up those compounds. One can learn of what DNA is made of through classic reductionism, but one can never learn the true function of what DNA really is with this process. The DNA molecule with all of its ramifications in the formation of living things is too complex to be understood from a purely reductionist perspective. How DNA fits into the natural order must be viewed wholistically. Truly understanding Nature requires that one looks at it in its entirety.
As a very mundane example of what we mean by this, assume one takes a picture of a Monarch butterfly with a new digital camera. If one then looks closely at the detector upon which the butterfly’s image has been captured, one would see it is made up of rows and columns of tiny pixels. Then if one further analyzes these rows and columns, one observes that each pixel behaves like a well that holds a discrete number of so-many electrons. Further isolation of the individual pixels reveals that the number of electrons vary from pixel to pixel like tiny steps, and each pixel contains a different charge. While this analysis leads to great understanding of charge coupling, where has the butterfly gone? It can only exist when the system is whole but is lost when the parts are taken away.
Complexity is an emerging new science and like Nature itself is a wholistic quantity. In order to show that it may be feasible for a man-made system – a system comprised of nanoelectronic components – to be brought to a state that can produce conditions compatible with those required for an emerging consciousness and subsequently sentience, the explanation of this system must be couched in the light of complexity and its wholistic nature. For example, everything in the universe is a physical entity and must obey the laws of physics. However, as we saw with the butterfly, not everything can exist after reduction to a simpler form. While biology obeys the laws of chemistry and chemistry obeys the laws of physics, biology does not reduce to chemistry and chemistry does not reduce to physics – contrary to the ideals of the String Theorists who believe they are on the verge of a Theory of Everything!
Computer Technology. So where are we going with this? A powerful driver for the development of nanoelectronics is for its application to computer technology. And to relate nanoelectronics to the onset of consciousness and sentience, we will need to discuss computers and what they are – and are not. Nature has spent several hundreds of millions of years in computer research and development and has converged upon a design quite unlike that which humans have focused on over but a few decades. The design selected by Nature is the neural computer, not a digital computer that we humans seem to have favored. Digital computers definitely have their place, as they are extremely precise. During the Cold War, for example, their precision was essential for the very accurate targeting required for ICBMs. And motivation for a computing system that could perform rapid arithmetic functions extremely accurately encouraged their continued development. However, research into neural computers has been carried along in parallel with that for digital computers, and in the decade of the Nineties it has suddenly accelerated. While neural computers are not as precise as digital computers, they are unequalled in parallel processing speed and are inordinately faster than digital computers for arriving at “approximate” solutions. While they lack the precision of digital computers, their results are often close enough. The classic traveling salesman problem that would take a digital computer 9.8 billion years at a million iterations per second to solve an optimum route for only 25 cities was recently handled by a research team using a neural computer in only eighteen months of computer time for 3038 cities! [160] Furthermore, a neural computer can modify its programming through inputs received from the environment, while a digital computer can not. Hence a neural computer is able to learn from its environment and modify its behavior as a result.
Consciousness and Sentience. The introduction of nanoelectronics and the demonstration of self-assembly of molecular wires and nanodevices into potentially millions of connections will enable extremely complex nanosystems. These nanosystems, because of their minute sizes, will be very susceptible to quantum effects. As more understanding evolves of quantum computing processes, [161-165] self-organizing structures will be designed that are receptive to the formation of coherent quantum states. The increasing knowledge and application of quantum computing, especially the development of self-organized quantum-coherence collapse times, will leave little doubt that some form of pre-consciousness or consciousness will naturally emerge and lead to sentience. Although inducing consciousness within a non-biological system will indeed be very difficult and complex, philosophers who assert that such an event is not possible underestimate the power of future technologies. And consciousness itself may not be as rare as we have come to believe in our andocentric world. [166-193] It is a very real physical phenomenon. While it still may be a mystery from the perspective of our overall lacking of its understanding, it still obeys all of the laws of physics. But it is a wholistic phenomenon and cannot be understood through physics in the reductionist sense. There is strong evidence that most – if not all – complex biological entities exhibit some degree of consciousness, regardless of species.
The objective of this thesis is not so much to “solve the problem” of how sentience will come into being as it is to “raise the issue” that its probable occurrence may be a natural progression. The goal is to suggest conditions – with sufficiently robust rationale – that were they to exist, then a form of sentience would emerge.
The intended path is to show how nanoelectronics applied to the development of neural computers – computers capable of adapting their behavior to the environment – can create within their nanostructures a condition of sufficient complexity to bring about the necessary “resonance” required of an emerging state of consciousness and sentience.