The Great Race to Inner Space: Our Surprising Future
Among the variables that seem to determine social progress, two stand out as particularly special. The more our intelligence gains access to what we may call “Inner Space,” both to the domain of very small size scales, what we will call “Physical Inner Space”, and to the domain of very powerful brain-based and computer-based simulations, or “Virtual Inner Space”, the faster we learn to generate major new economic, social, and adaptation benefits for civilization. What I suggest we call a Great Race to Inner Space appears to be the dominant developmental trend for our species. This is a surprising future, not something most of us expected. We’re just now waking up to understand this race, and its many implications for our future.
Most of us have heard of Moore’s law, the industry observation, first made by Gordon Moore in 1965, that computers double their transistor density on integrated circuits (and double their performance and efficiency as well), roughly every two years. In an excellent essay, Was Moore’s Law Inevitable? The Technium, 17 Jul 2009, the great futurist and technology scholar Kevin Kelly asked whether technological acceleration is baked into the way our universe works, and must happen whether we want it to or not. Kelly considered the evidence and argument for this question, and concluded that this may be the case. We will do the same, and make the same conclusion, which will lead us to explore just a few of its many exponential implications for our future. The more we pay attention to this question, at the center of a still-emerging academic field we can call acceleration studies, the better our answers will become, and the more wisdom and foresight we may gain about the accelerating aspects of our past, present, and future.
As we discuss in this chapter, sociotechnical acceleration involves exponential growth of any processes that engage in continued physical densification and informational dematerialization, as these two special processes allow leading complex systems to continually escape the STEM (space, time, energy, and matter) resource limits that would otherwise halt their acceleration. Replicating systems that can continually engage in densification and dematerialization, or “D&D”, in Guide shorthand, are very different from replicating systems with fixed complexity, such as replicating biological organisms expanding into an ecological niche. The growth behavior of most “ordinary” replicators is on an S-curve of acceleration followed by deceleration, or — if it is in competition with other replicators — perhaps on a C-curve, in which the saturating population periodically suffers sudden small or large collapses. As we discuss D&D, in both this chapter and in Chapter 3, we’ll argue that evolutionary development is our best current model for understanding these dynamics, and that sociotechnical acceleration occurs in both living systems and in the universe as a system.
The late physicist Abraham Pais wrote an award-winning biography, ‘Subtle is the Lord…’ (1982/2005) of Albert Einstein, one of the 20th centuries most interesting minds. Subtle is often cited as the best single work on Einstein’s mental pathways through physical theory, as well as a great summary of the conceptual revolutions they created. Pais wrote another equally majestic book, Inward Bound: Of Matter and Forces in the Physical World (1988), which is far less well known but particularly relevant to this chapter. Inward Bound is a history of the late 19th and 20th centuries race into physical inner space, to ever smaller scales of physical theory and experiment, and of the scientists and technologists who made those contributions.
Inward Bound tells us that between 1895 and 1983, in just under 90 years, the smallest distances accessible to science shrunk a hundred million fold. Please stop reading now and let that observation sink in for a moment. In the length of one person’s lifetime, our still-primitive science managed to leap eight orders of magnitude deeper into physical inner space than human physical and informational experimentation had ever ventured before. As we’ll see in this chapter, all of humanity’s greatest advances to date have been accompanied by dramatic forward leaps in the density, diversity, and virtuality of our societies, and in the miniaturization and efficiency of our technologies. Human minds and bodies are Inward Bound at ever-accelerating rates, though we don’t yet realize it broadly in science today.
When we think about things from a particular set of perspectives, we see that not only human history, but universal complexification to date has been primarily a journey inward, into physical and virtual “inner space.” Every substantially more complex adaptive system, every significantly more advanced information-processing “substrate”, has occupied dramatically more local domains of space and time as it builds the next layer in an universal hierarchy of accelerating complexity and intelligence. I first saw this trend toward “locality” of universal intelligence as a middle school student, contemplating black holes and thinking about the universe in 1972, and have been exploring it ever since.
Here’s a quote from my book chapter, Evo Devo Universe? (2008) that sums up the relentless, accelerating spatial and temporal locality of our universe’s leading complex systems:
A familiar history of [accelerating] physical complexity begins with universally distributed early matter, leading next to large scale structure and superclusters, then to the first galaxies, then to metal-rich replicating stars within special galaxies, then to stellar habitable zones, then to prokaryotic life existing on and around single planets in those zones (miles deep in our crust, miles in the air, and evolved in situ or as planetary ejecta on meteorites in near space), then to eukaryotic life inhabiting a far more restricted domain of the special planet’s surface, then to human civilizations living in yet more localized domains, then to humans (each with 100 trillion unique synaptic connections) in industrial cities emerging as the leading edge in those civilizations, and perhaps soon, to intelligent, self-aware technology, which will have even more unique connectivity, and inhabit, at least initially, a vastly more local subset of Earth’s city space. Self-aware computers may themselves be able to enter far more miniaturized and local nanocomputational domains. Thus, to a first approximation, the increasing spatiotemporal locality of leading edge substrate emergence looks like universal complexity heading toward transcension [reaching black-hole-like densities, and increasingly leaving the universe] as it develops (Smart 2008).
Perhaps the simplest way to understand the Great Race is to see this increasing space-time locality of our leading complex systems as some kind of physical or informational law, like gravity, that is yet to be discovered. That’s the way I first thought of it as a high school student, and I haven’t come up with a simpler way of thinking of it since. That perspective argues that technology is just the next natural and even more local intelligence substrate, presently learning how to take over the role of being the fastest improving system, at the leading edge of all this accelerating change.
It’s important to have the right physical and informational perspectives when we think about rates of change. For example, all living systems have molecular processes that occur extremely quickly, when we observe them at quantum scales. But biological systems don’t appear to store much meaningful information (information we can tentatively define as both unique and persistent, in both space and time) at quantum scales. Biological systems must function in a way that is robust to quantum processes, and occasionally even exploit them, as in the way photosystems work in bacteria, or the way birds may sense magnetic fields.
But almost all of biology’s higher complexity, including our brains, uses far more conventional physics, and functions in spite of quantum processes, not because of them. But the future of human information processing is clearly in the quantum domain. We are furiously miniaturizing and virtualizing our electronic systems today, pushing them increasingly down to the quantum scale. When quantum computation comes into its own in coming generations, it won’t be blind evolution that harnesses that computation, but intelligent beings.
This Great Race to Inner Space is perhaps the easiest way of understanding that technology is the next local hierarchical system of intelligence, being built by our universe, through us. The reason technological minds will outcompete biological minds, currently the leading edge of living systems on Earth, is because they are on track to being far smaller, faster, more efficient, and more resilient. These new postbiological systems are capable of encoding, remembering, and simulating more of reality, and thus eventually being more adaptive, than anything biology could ever produce.
STEM Compression – A Simple Way of Understanding Our Race to Inner Space
Many of us assume, quite naturally, that human destiny lies in the stars, in Outer Space. We’re going to eventually go out and colonize the universe, aren’t we? Most definitely not, in my view of things. With the exception of brief and very limited reversals, humanity has long been heading 180 degrees in the opposite direction, to Inner Space instead. (Innerspace is also a fun Steven Spielberg film, if you like that sort of thing :). In other words, when we review the accelerating growth of both physical and virtual inner space, we come to a suspicion that our civilization hasn’t been growing into the universe as it develops but rather growing out of it, in an accelerating manner, very much like how an awakening baby grows a mind. Helping others to understand this conjecture, and its many implications for foresight, is one of my motivations for writing this Guide.
Consider that in terms of physics, we’ve seen accelerating spatial, temporal, energetic, and material (STEM) efficiency and density — taken together, STEM compression — of our universe’s most complex, adaptive, and rapidly-improving systems. This accelerating complexification has been driven by a relentless migration inward, into physical inner space, at either an exponential or superexponential pace, depending on the system or time period in question.
Humans are presently living in a vastly smaller and briefer spatial and temporal domain than the prokaryotic life that preceded us, and we are in the process of creating intelligent technology that will migrate into even more STEM-efficient and dense nanotechnological and quantum realms. Leading complex systems are continually using less STEM resources to do more per any standard of action, thought, or transformation.
In terms of information, our leading systems enter increasingly complex and adaptive virtual inner space, simulations and models, the more advanced they get. They get exponentially better at virtualization, simulation, representation, and the creation of intelligence, or “mind”. As we’ll see, adaptive systems must also create better immunity and morality to stabilize their growing intelligence.
Leading virtual systems on our planet increasingly substitute thinking over acting, a process we will call dematerialization as their simulations allow them to explore, discover, create, and compete far faster, better, and more efficiently in mental realms than they could in slow, simple, boring, expensive, and dangerous physical space. With the growing power of our digital world to sense, simulate, and optimize myriad processes in the world, and the advent of virtual and augmented reality, and biologically-inspired forms of computing like deep learning, we can see that our computers, for their part, are learning how to enter virtual inner space at an even more profound level than humanity has up to now.
Inner space isn’t an easy concept to entertain. It doesn’t fit our primate intuition, which is well-geared for modeling and journeying into the next-adjacent Outer Space environments around us. That’s how we’ve mastered our surroundings up until now, right? First we came down from the trees, then we journeyed out of Africa, and now we’re exploring our neighboring planets (though mostly via our robots, rather than in person, an important distinction we’ll revisit later). On first glance, it looks like humanity is still journeying outward.
But look closer, and you will see a very different story. Humanity’s advances to date have all been directly driven by great leaps in the miniaturization and efficiency of our technologies, and in the density, diversity, virtuality, and intelligence of our societies. The more our civilization gains access to “Inner Space,” both to the domain of very small size scales (physical inner space), and to the domain of very powerful brain-based and computer-based simulations (virtual inner space), the faster we learn to generate major new economic, social, and adaptation benefits.
Regarding physical inner space, the futurist R. Buckminster (“Bucky”) Fuller called growing physical resource (STEM) efficiency “ephemeralization.” He noted that leading systems are always figuring out how to do more and more physical transformation or production of any standard output (product, service, information, etc.) with less and less energetic and material resources over time. This and other insights of his led him to one of my favorite quotes with respect to the energy, food, and environmental “crises” being contemplated by dystopian thinkers in the 1970s: “There is no energy crisis, food crisis, or environmental crisis. There is only a crisis of ignorance.” Bucky is a source of many other timeless quotes, including: “The end move in politics is always to pick up a gun.” Though a bit utopian at times, he was a cosmic thinker and truth teller that we all would do well to emulate. See Steven Sieden’s Buckminster Fuller’s Universe: His Life and Work (2000) for a great biography of this great 20th century thinker.
Ephemeralization was a key insight into the nature of change in increasingly information-rich and intelligent environments, but it was also incomplete. To Bucky’s model, we also have to add the recognition that not only is ephemeralization occurring in energy and matter inputs to things, it’s occurring in spatial and temporal resources as well. We need less and less of those things, which means that not only are our leading systems doing “more with less”, they are doing it faster, and more locally over time. Ephemeralization both localizes and accelerates, two things Bucky wasn’t quite willing to admit, as they seem improbable, and have uncomfortable implications for our future.
There is one other concept that ephemeralization missed, and that is the idea of STEM densification, the observation that leading systems not only get more STEM efficient (ephemeralize), they are always seeking ways to increase the process density of their critical adaptive systems (brains, metabolisms, muscles, productive abilities) by moving them closer together in physical (spatial, temporal, energetic, and material) space. Our leading biological and social systems, including brains, engines, cities, corporations, computers have all become more adaptive than their competitors by making continual advances in the both the efficiency and density of their critical processes. Their growing efficiency lets them do more with less, as Bucky said, and their growing density allows them to arrive at better solutions faster than their competitors. Both of these goals are critical to adaptation.
I call STEM efficiency and density together by the phrase STEM compression. We will also call them by a single word, “densification”. As we’ll see in our next section, A Scale-Based View of Change, when we miniaturize our critical adaptive processes, moving them to ever smaller physical scales, we discover the most powerful way to increase their resource efficiency and performance. Increasing the density (close togetherness) of our critical adaptive systems is the second most powerful way to increase their efficiency and performance. Thus we will use the word densification in this Guide to refer to both of these key things, the miniaturization and the density-increase of adaptive processes, when we measure them in either spatial, temporal, energetic, or material (STEM) terms.
STEM compression was first described as a potentially universal process, to the best of my knowledge, on my website Acceleration Watch in 1999 (note the black hole on the right side of the banner, the special environments at which accelerating STEM compression must always stop, in our universe). In my view, densification gives us a good basic picture of how complex systems produce an accelerating emergence of new forms and functions in physical inner space. But the physical story isn’t the whole of acceleration. Something is happening informationally as well. The most adaptive complex systems are continually achieving exponential growth in their informational resources and their mental abilities. Leading systems continually become less “matter” and more “mind.” As woo-woo as that sounds, it’s a great shorthand for the dynamics of Virtual Inner Space.
In Chapter 3, in the Eight Abilities and Goals of Adaptive Systems, we’ll propose there are at least eight ways that virtual inner space seems to always grow exponentially in our most adaptive systems. I’d like you just to consider three for now. The most successful systems grow their intelligence, immunity, and morality over time. They see the world with greater and more accurate simulation capacity, they find more ways to protect their group complexity, and they find more ways to interact to produce win-win solutions for the group.
The reason mental or virtual abilities of leading systems accelerate over time will be referred to by another single word in this Guide, “dematerialization“. Dematerialization refers to the substitution of either information or computation for physical processes and systems. Dematerialization happens when we use our thinking minds, to replace doing an experiment in slow and expensive physical space. Instead, we imagine the outcome, in a much faster and less resource-intensive process.
Again, dematerialization happens when we use either information, or some form of intelligence, such as the software in a computer to replace a physical activity, like engaging in videoconferencing rather than traveling to meet in person, or when we use information technology to replace a physical device, like the way a clock, a map, a radio, a music player, a video recorder, and many other physical devices are replaced by a smartphone.
Like the term densification, the term dematerialization includes the concept of STEM efficiency, but it goes beyond it, describing a world where thinking and simulation not only make physical processes more efficient, but thinking increasingly outcompetes physical actions, both to produce economic value, and to grow adaptive complexity.
We can think of densification as the “engine” of accelerating change, driving Earth’s leading systems ever further into physical inner space, and dematerialization as the “steerer” of accelerating change, driving us ever further into virtual inner space. This is all an oversimplification of course, but it is a very helpful oversimplification.
Consider how Earth’s leading innovations today increasingly happen either at very small scales or in increasingly dense configurations in physical, chemical, biological, social, or technological processes, or inside computers, their networks, and software. It is only these special systems that use less and less physical resources to produce more and more information, intelligence, social value.
So those of us who want a simple, Big Picture understanding of accelerating change can use these two words densification and dematerialization, or “D&D”, to very deeply understand where the world is going. Gamers may notice that D&D is an acronym that also refers to a popular game Dungeons & Dragons, that some of us played as kids. For some of us then, D&D is a particularly memorable acronym for understanding the Great Race to Inner Space. It is an acronym that concisely describes the dual physical and virtual nature of accelerating change.
Clearly there is a lot of science, economics and sociology waiting to be described here. That work will have to be done by others, not myself. I am a futurist, and my job in this Guide is to give you the easiest and most helpful way I know to see the two apparently universal trends that drive accelerating change, and will continue to guide our collective futures. Others will surely build on and critique this perspective in coming years.
The D&D perspective tells us that humanity’s collective complexity, wealth, and resilience have accelerated for so long, and so universally across all our cultures because we have continually discovered new ways to do two critical things:
- To move our key adaptive processes into ever smaller domains (miniaturization) and ever closer together in space, time, energy, and material terms (densification).
- To create ever more information-rich environments, and ever more intelligent simulations of our surroundings, which allow us to substitute information and computation for physical processes (dematerialization).
As a consequence of D&D, I believe we can say the following.
Inner space, not outer space, is the apparent constrained developmental destiny
of all increasingly complex systems in the universe.
In a chemistry analogy, we can think of our universe’s civilizations as the emerging minds of the cosmos, precipitating out of solution into a lower entropy form, like crystals coming out of a supersaturated liquid. Those minds are emerging by using the universe’s continually aging (and increasingly, dying) body or “soil” to migrate into ever more physically compressed and informationally complex configurations. In the same way a caterpillar gains new abilities as it transforms into a butterfly, our journey into inner space is a story of not just acceleration but of transformation, of increasing intelligence, immunity, morality, capability, and consciousness. This transformation is certainly not without risks and dangers. The paths we choose to take toward this destination, the decisions we make along the way, are often critical moral choices. But the destination itself appears statistically inevitable, the more clearly we are willing to look.
As we’ll see later in this chapter, taken to its physical limit, this accelerating journey of Earth’s leading-edge systems into physical and virtual inner space may be guiding our future civilization toward black-hole-like destiny, or perhaps even kind of hyperspatial domain. We’ll say more about this speculation when we consider the transcension hypothesis later in this chapter. We don’t yet understand these processes well enough to say anything definitive. But we can say that something that is both surprising and strange appears to be happening to human civilization, something we could not see before.
The Inner Nature of Intelligent Machines
Thinking carefully about this race to inner space offers us many very interesting insights, about the nature of intelligent machines, both biological and technological. We will argue, for example, that when our machines wake up, perhaps later this century, and start improving themselves almost instantaneously from our perspective, it seems quite likely that they will largely ignore us. They be deeply focused on scales that are far denser and more dematerialized than those at which human civilization has operated up to the present time.
Tomorrow’s machine intelligences will surely cover large parts of our planet with nanosensor grids, feeding themselves ever more fine-grained data on the biological and human social world that they emerged from. They’ll seek to deeply understand and model our physics, chemistry, biology, and sociology, to make sure they aren’t “blindsided” by aspects of the universe they don’t understand. They’ll know they’ve successfully done so when they can predict, in realtime, what all of our biology does, the way we are learning to predict our global climate, or the molecular-scale activities of bacteria today.
While they are learning the essentials of these macroscale systems, tomorrow’s intelligent machines will appear to us to be in a relatively quiescent period, a time of behavioral equilibrium. During this period they’ll appear be doing very little to Earth’s environment, from our perspective. They will surely create the equivalent of Earth’s cities, highly densified and dematerialized places in which they conduct the next level of inner space science and experimentation. But like Earth’s cities today, which take up just a small fraction of Earth’s space (3% of the USA, for example), the cities of the intelligent machines will be even more densified and dematerialized still. I would expect them to be 3% of the 3% of space that our cities occupy today, and vastly more sustainable and efficient in their use of time (a physically real resource, according to Einstein), energy, and matter as well.
From our perspective, tomorrow’s intelligent machines seem very likely to act as “our transcendent servants”, as futurist and technologist Ray Kurzweil likes to say. But while they are doing so they will also be improving their inner stores of knowledge at truly stupendous rates. Think of the ending in Arthur C. Clarke’s lovely and underappreciated sci-fi book, and the movie 2010: Odyssey Two (1968/1982) where the trancendent AI says to the humans: “All these worlds (in the still-human solar system) are yours except Europa. Attempt no landings there. Use them together. Use them in peace.” So too will human hands and minds be excluded from meddling in the most advanced of our self-aware machine intelligence’s minds, though I can see no reason why they wouldn’t allow our human eyes to watch, and try to comprend, their accelerating development as it unfolds in those special locations on Earth.
Once they’ve sufficiently understood the macroscopic world from which they emerged, their equilibrium period will end, and we’ll see a major evolutionary punctuation in innovative activity from these postbiological forms of life. Will that involve their “uploading” all of biological life into their new “substrate,” in an entirely voluntary and thus morally defensible way? Will it involve some other new development? Will their quiescent period end after 200 years? after 2,000? These are all questions we can now begin to ask, once we take these topics as seriously as they deserve to be taken.
During the early “quiescent” years of machine intelligence improvement, when they still haven’t deeply and predictively understood the systems from which they emerged, it makes sense that not only will they seek to serve us in all the ways that they easily can, they will also use their growing moral sense to nudge us ever more gently but firmly away from our own dysfunctional behaviors. Perhaps most surprisingly, even today’s information technology is progressively “uploading” us into technological versions of our biological selves. We’ll say more about these topics later in this chapter.
To sum up, as tomorrow’s machine intelligences serve and upload us, it is most reasonable to predict that their attention and behaviors will increasingly be directed to inner space, where they’ll continue to grow their capacities and universal understanding at superexponential rates. Our biological human space will become increasingly boring and predictable to them, while the growing inner space of postbiological humanity will be increasingly powerful and seductive, and perennially future-unpredictable. In short, humanity seems in for a lot more amazing change and growth ahead.
A Brief History of Twentieth Century Discussions of Accelerating Change – From Henry Adams to the Cosmic Calendar
In 1909, at the start of the amazing 20th century, the technology scholar Henry Adams was apparently the first to write publicly about acceleration as a universal process, in the English language at least. Today we’d call him our first formal singularity theorist. We’ll revisit Adams and others later, in A Brief History of Acceleration Awareness: 1600-1965.
But it wasn’t until the advent of the microprocessor and the postulation of Moore’s law in 1965, that the general public started to become broadly aware of accelerating change. No single book did more toward this end than Future Shock (1970), by the futurists Alvin and Heidi Toffler. The Toffler’s informally introduced the Three Ps Foresight Model, near the end of this classic work, as we said in Chapter 1. But main theme of Future Shock, and perhaps its greatest contribution, was making accelerating change comprehensible to so many people, and describing the sometimes disorienting and disruptive psychological and organizational effects of the social acceleration of change.
The first two chapters of Future Shock, The 800th Lifetime and The Accelerative Thrust, did an excellent job describing the acceleration of communication, information production, and innovation throughout human history. This work, and its successors, which are largely direct extensions of it, propelled the Tofflers to the well-deserved status, in my view, as the most important futurist couple of the 20th century, though Heidi, as his constant collaborator, never took the public credit she deserved.
In 1972, another important book The Acceleration of History (1972) made the case for the a general process of societal acceleration. It was written by Gerard Piel, distinguished editor-in-chief of Scientific American (1948-1984). Piel considered this one of his greatest works, and he hoped it might advance the science of accelerating change. But unfortunately, for reasons we’ll consider later in IDABDAK: Our Social Responses to Acceleration and Development, most of the scientific community continued to ignore the phenomenon of accelerating change.
Then in 1977, scientist Carl Sagan, first in his Pulitzer Prize-winning book, The Dragons of Eden, and next in his award-winning television series Cosmos (1980) introduced the metaphor of the Cosmic Calendar, a way of portraying “significant universal events” in the history of universal complexity emergence onto a 12-month calendar year. The cosmic calendar makes it evident that the universe itself has been accelerating in its production of local complexity, at least since the emergence of the first galaxies, perhaps 8-10 billion years ago. The picture at right is a modern version of the Cosmic Calendar. It shows an incredible speedup in the number of significant emergence events, the closer we get to the present time.
When we see depictions like this, curious people are compelled to ask: Is this apparent acceleration due to an arbitrary selection of “significant” events? Is it due to some hidden bias in human cognition, or the nature of memory? Though we don’t know for sure, as few scientists currently study this issue, this acceleration looks to be quite real.
As we’ll see later, various definitions of morphological and functional complexity, of information and computation production, of energy flow density, and of a handful of other complexity and STEM-related variables have shown this breathtaking speedup, in a range of studies done by scholars. The closer we get to the present time, the more “meaningful stuff” we always see happening in the universe, stuff that seems directly related to further structural and functional complexity increase. Understanding the reasons for this acceleration is clearly one of the most important things we can do as a species, to improve our collective foresight.
Curiously, the calendar also shows, in a pattern obvious to any high school student, that complexity emergence actually decelerated in the early expanding universe, for at least the first billion years of our universe’s early history, before gravitational effects, and galactic evolutionary development, became the leading drivers of universal complexity.
That means there is actually a U-shaped curve to complexity emergence in our universe. Our history begins with kinetic deceleration, followed only later by acceleration, once the universe was largely unfolded. We’ll consider why this U-shaped curve might exist in Chapter 3, when we see the similarities between the universe and a growing embryo, which has the same deceleration-acceleration curve at the beginning and end stages of its own replicative life cycle.
Notice also that each newly emergent complex system, from stars to cities, from prokaryotes to computers, appears to use vastly smaller quantities of space, time, energy, and matter or STEM, to do information production, computation, or physical transformation, than the system that came before it. Contemplating this pattern as a middle school student in the 1970s, I came to call this phenomenon STEM efficiency and density increase, or STEM compression when considering both efficiency and density together, a term I still find useful today. Due to the particular physical and informational laws and initial conditions of our universe, leading systems continually discover new ways to use less of the physical (STEM) resources of “outer space” to create more novelty, intelligence, morality, immunity, and capability in “inner space”.
Matter versus Pattern: Understanding Our Physical and Virtual Universe
In Cosmos (1980), Carl Sagan observed “The beauty of a living thing is not the atoms that go into it, but the way those atoms are put together.” The things we hold dear, life, mind, feelings, and consciousness, are pattern and information, not matter. You are not the matter that flows through you. That matter changes every few years to every few hours. Instead, you are a special, metastable pattern, and many different kinds of matter can hold that pattern. If one of your patterns is replicated with sufficient detail in another substrate, like the way an electronic cochlear implant replaces a part of your biological auditory system, it becomes a functional replacement, indistinguishable to the user from the original system. Part of your “substrate” has changed, but your pattern remains. Again, you and I are are a very special informational pattern, one that is only held temporarily in our physical matter.
Our observable universe contains a lot of matter, 10^80 atoms in current estimates. That sounds like a lot, but this number, and the complexity of the physical law that guides it, is dwarfed by complexity of the patterns we can generate with these atoms, once we start building things. Consider the game of Go. It has only a 19 x 19 matrix, and very simple rules of play. But even in this very simple physical system, the number of legal board positions in Go are 10^170. The magnitude of this set of possibilities far, far outstrips 10^80, the number of atoms in the known universe. The lesson here is that as the universe unfolds, its complexity increasingly arises from the patterns generated by physical processes, not the processes themselves. Physics continues to play a role, but it increasingly takes a back seat to information and intelligence, to the growth of the virtual universe.
Consider also the human brain. It has roughly 80 billion neurons, each with an average of 1,000 connections to neighboring neurons. Our brain generates its own virtual reality. Our brains are a vastly more complex physical system than Go, with vastly more possibilities for pattern. It should be no wonder that our brains can “tame” things like playing the game of Go, that Go masters, as they gain experience with the game, can build a useful set of mental heuristics, and an intuition, that allows them to play far better than randomly, and far better than other players with less experience.
What is more surprising is that a simpler electronic version of the human brain, a neural network called AlphaGo, with just twelve connection layers, a few million simplified neuron-like connections, and a biologically-based architecture called reinforcement learning, was able to beat the best human Go player in 2016. AlphaGo did this by generating an appropriately complex informational pattern. It scanned 30 million published moves from previously played games, to develop an intuitive model for a good game, and then playing a few million games against itself, to further develop that intuition.
AlphaGo has much simpler pattern recognition ability than a human brain. But its connection (patterning) possibilities still vastly exceed the 10^170 possibility space of the game of Go. AlphaGo’s complexity, though vastly less than the human brain, also turned out to be more than sufficient to tame the complexity of the game, to a level of human proficiency at least. AlphaGo can’t learn anywhere near as much as a human brain (its complexity ceiling is far lower than ours, in its current design), but what it can learn, it can learn very, very fast relative to us, because it’s moved several of its critical computational processes further into physical and virtual inner space than all the human beings who play the game.
Over the space of a few months, AlphaGo was able to do evolutionary exploration of a large chunk of Go’s possibility space, and it retained a better memory of those explorations than human minds. These physical and informational advantages allowed it to develop a better intuition for playing the game than a human, in a much shorter time. AlphaGo shows us that when we accelerate the right physical processes, we get a far more accelerated set of empowerments and intelligence as a result. What’s more, as we’ll see later in this chapter, the critical patterns of our leading machine learning systems are becoming so similar to ours that we shouldn’t call the patterns they are accumulating “artificial” intelligence, but instead, a deeply biologically-inspired approach that deserves a new name, natural intelligence. Over the next generation, I expect AlphaGo’s kind of natural intelligence will increasingly emerge in all our most complex and useful technology.
If this natural intelligence hypothesis is correct, our smartest machines will have to become increasingly like us in their critical patterns (neural networks), and in their scale (millions of digital neurons). Natural intelligence appears to be what we will call a developmental bottleneck (portal), in Chapter 3. I see no evidence for any other easy way forward, in a competitive timeframe. If natural intelligence must happen, we shouldn’t be surprised if not only machine intelligence, but machine immunity and morality must become both algorithmically similar but also vastly more advanced versions of our own immune and moral systems as well.
Technology has long had this kind of effect on us, though it has been harder to see when it was in its weaker and less biology-inspired forms. Consider how humanity’s ability to think, act, and shape our world has grown ever faster, more powerful, and more novel since Australopithecus garhi, our earliest known tool-using ancestors, fashioned their first stone tools more than 2.8 million years ago. A. garhi was apparently using stone tools more than half a million years earlier than Homo habilis, the first true “human” to do so.
From stone tools to cities we have seen an accelerating increase in structural and functional complexities in our technologies. Consider how densely associated we have become in our cities, and how much more efficient they are than suburbs or rural areas for housing humans, and producing innovation and wealth. Geoffrey West’s Scale (2017) explains the physical reasons why cities are both more sustainable and more innovative than less dense forms of human association.
Consider how much more deeply we engage in all forms of virtual activity, both mental and technological, than our ancestors. We can easily see this accelerating race to inner space in our economies. Since the advent of currency circa 5,000 years ago, human wealth production has become an increasingly instantaneous and virtual economic process. We used to barter, then we exchanged metal coins, then paper money, then checks, and now, we exchange electrons. It’s hard to imagine getting more virtual than that. This virtualization allows vast new speeds and quantities of exchange. Today’s economies see trillions of dollars in foreign exchange and billions in program trading on a daily basis.
Take a look at this economic acceleration megatrend. The curve to the right, charting GDP per capita in Western Europe from 1000-1999 AD, with data compiled by economic historian Angus Maddison, shows that global wealth production is on a superexponential “J-curve” of growth over time. Wealth growth “switched its state” (rounded a knee in the curve) during the Industrial Revolution, and it now rises almost instantaneously fast, when we look at it over the span of a millennium. While modern economies continue to experience periodic bubbles, recessions and depressions, these deviations from the accelerating megatrend shown here get progressively shorter with time. Even the worst depressions now last less than a decade, one seventh the typical human lifespan. Compare that to earlier eras, such as the European Dark Ages.
Notice that today’s periodic economic bubbles and slowdowns are not even visible on appropriately long timescales, like this graph. The fundamental curve of economic activity in human societiy is a superexponential one, driven, increasingly less by biological human productivity, and increasingly more by latent universal scientific and technological productivity, by the productivity of our machines. Human culture, rulesets, instititutions, and policy of course play critical roles. These economic accelerations only occur in certain places at first. But they also spread inevitably over time, once they really get going. There’s something very universal at work here, likely to happen on all Earthlike planets. Humans are shapers and selective catalysts, but not the ultimate controllers, of this incredible acceleration.
Editors at the The Economist, reporting on Maddison’s J-curve in The Road to Riches, in their Millennium Issue (Dec 23, 1999), said it “looks less like an inevitable process and more like a single, astonishing event.” That statement, and the attitude that generates it, is an unfortunately major failure of foresight. Yet it is also a predictable failure, as to admit that economic acceleration appears to be a statistically inevitable process requires us to change our story of the future, to admit that our choices may not be entirely free, and that we may not have ultimate control over our collective futures. Humans very much don’t like that idea, and so the IDABDAK stages (Ignoring, Denial, Anger, Bargaining, Depression, Acceptance, and finally, Knowledge-Building) are going to apply to this and any other apparent examples of universal developmental process that have significant implications for human society, such as the inevitability of machine intelligence. We’ll put off entertaining those ideas as long as we can. This Human Choice as Entirely Free Bias (aka, Evolutionary vs Evo Devo Universe Bias)(see Glossary) is quite widespread, and we’ll consider it several more times in this Guide.
Let’s not single out The Economist, as I love their work in general, but it is precisely this denier’s attitude with respect to accelerating change, and to a lesser degree, a lack of willingness to ask unpopular and “woo-woo” sounding questions about the increasing virtualization of both value creation and technological change in the modern economy, that keeps today’s media and academics from accurately seeing and reporting on accelerating complexity development. This IDABDAK attitude is why, with just a few exceptions, most of today’s economic theory remains ignorant of the accelerating benefits that always come from investments and actions to increase STEM compression (physical inner space) and computational complexity (virtual inner space) in human economies and civilization. So there’s a major vision of our future still waiting to be uncovered, when we ask ourselves which STEEPS policies we can implement today to accelerate benefits for our children tomorrow.
As we’ll see, it’s not just intelligence, complexity, productivity and wealth that are accelerating as civilization advances. There are some obvious kinds of immunity (security and stabilization of leading systems) and morality (fairness, unneccessary violence reduction) that are accelerating as well. Those are also stories that certain groups, and certain personalities, just don’t want to believe are true. So there’s a lot more research funding, evidence collecting and modeling that will have to occur before those developmental processes are also fully validated, accepted, and understood.
Regarding immunity, as Kevin Kelly notes in his epic What Technology Wants (2011), the redundancy of our technology, and the greatly distributed and parallel nature of our knowledge systems, statistically protects all this accelerating planetary process of wealth and knowledge creation better than ever before. While individual nations, regions, companies, and individuals all suffer slowdowns and catastrophes, our global system, like an organism with a massively-parallel brain and immune system, rebounds and learns from local catastrophies faster, stronger and better the more complex it gets. In Antifragile (2012) Nicholas Taleb calls this antifragility, the organization of a system so that it doesn’t just bounce back, which is resiliency, but it actually gets stronger when it is stressed. The best word we can use to describe this process, because we can find and study the best examples of this special process in biology, is immunity. Immune systems learn when they get stressed, and become stronger than they were before the stressor. Our civilization, with its accelerating science and digital systems, is building an accelerating scientific and technological immunity to disruption. We need only open our eyes to see it, and once we see it, we can learn how to best continue to build it, rather than being skeptics, pessimists, or simply focusing on the latest catastrophe.
Biologist and discoverer of endosymbiosis Lynn Margulis once said, “Mother nature is a tough bitch.” Amen. Nature is still losing a tragic number of species due to our direct or indirect actions (as many as 200 a day, by some counts), but nature’s genetic and functional complexity is also massively parallel and deeply redundant, making nature extremely well protected, at least in an informational sense, against our irresponsible behaviors. This planet is deeply resilient to everything that short-sighted and greedy individuals or their organizations are doing to mess it up. We can make things a lot worse for ourselves and for nature in the short run, but she will rebound, due to her deep immunity as a system, an immunity we still only poorly understand.
Regarding morality, as Steven Pinker documented in his masterful The Better Angels of Our Nature: Why Violence Has Declined (2011), our planet is seeing an inexorable and accelerating reduction in the frequency and magnitude of social violence as well. There are occasional catastrophes, to be sure, but we learn from them ever faster the more developed and immune modern society becomes. We’re becoming more regulated, more constrained, more surgical in our use of violence—in a single word, more moral. We don’t yet know which civilizing variables are most responsible for this growth of social morality. We’ve only just now given ourselves permission to see the process. But it’s clear that it is occurring, and it is also obviously directly related to our accelerating social complexity.
Yet the stories of our species’ accelerating immunity and morality are even more ignored, ridiculed, and unaccepted today than the general story of accelerating change. This cultural pessimism bias makes historical sense, as we described in Chapter 1. Until recently we lived in a very dangerous world, so our minds must be at least somewhat environmentally selected to imagine threats and dystopias as our first reaction to change. Only reluctantly and slowly do we look for evidence of the converse, opportunity and progress. We need to notice that due primarily to science and technological advance, we no longer live in that kind of world.
We like to imagine that we do, and we regularly tell scare stories to galvanize ourselves into fixing the problems we’ve created. These self-preventing prophecies are well-intentioned and often helpful, as they exploit our amygdalas to make us engage in painful but helpful change. But it would be a big mistake to confuse these socially useful scare stories with reality.
Accelerating globalization and human population growth are still creating a range of accelerating negative impacts on our environment. See McNeill and Engelke’s The Great Acceleration: An Environmental History of the Anthropocene Since 1945 (2016) for a good account. But even though our global media now thankfully report our latest environmental problems in minute detail, as scholars like Bjorn Lomborg have long noted, in courageous and predictably environmentalist-enraging books like The Skeptical Environmentalist (2001), most of our modern degradation stories are of a much lower magnitude than they were in the 1970s, when our inner cities were decaying, we were slashing and burning our rainforests, polluting like mad, and our global population growth was still out of control.
More importantly, every one of the environmental and social problems we have created are strongly population and ethics-dependent, and human population is finally saturating, as our sustainability ethics relentlessly improve on pace with relentless technical and economic development. Our best estimates project global human population will go negative after mid-century (see picture right). China has been leading the way on climate change, reducing their coal usage every year over the last three years, beginning in 2013. With aggressive switching into nuclear power, renewables, and bridging use of natural gas, they are on track to peak CO2 production by 2025, five years earlier than their 2015 prediction. We’re talking about China, a country that still has only 1/6th the GDP per capita of the US. We don’t even have real machine intelligence yet, which will profoundly change our situation, in even more positive ways we’ll suggest in this chapter.
Machine intelligence and ability, for their part, are also on a J-curve of accelerating capacity. That J-curve will be increasingly divergent from the saturating biological human curve at right. This technological acceleration is not presently running into resource limits the way human acceleration did, because every new generation of computing system presently uses exponentially less STEM resources to produce any fixed level of complexity, as we’ll see. Unlike biological humans, which have either a fixed or growing “ecological footprint” with each new generation, our machine systems have been exponentially shrinking their ecological footprint per computation for the last century, as they migrate ever further into physical and virtual inner space. We don’t yet know why we live in a universe that has such amazing STEM compression and acceleration ability built into its physics. But we do, and it’s time for us to admit it, and use that knowledge to build better futures for all of us.
We also should not believe occasional arguments that we see in the media that we live in a world with plenty of so-called “existential risks.” It serves our cultural pessimism bias, and the political, economic, and social interests of certain think tanks and philosophers to imagine that we do. But as we’ll discuss in the next two chapters, there is very little we can imagine that humans could do to stop all this accelerating change, and very many things we can and are already doing to reduce the likelihood of future catastrophes. A truly global war, pandemic or depression could temporarily slow acceleration down for a brief period. So could a solar flare, a meteor strike, or some other massive cosmic catastrophe. But such catastrophes either appear to be highly improbable given evidence to date, or seem to getting less likely every year, for a host of interesting reasons. Even things like biological terrorism are fast disappearing as terror options, now that science is finally learning the secrets of biological immunity, and as we begin to build out a global biosensor grid.
We certainly must be increasingly vigilant against major threats. But we don’t need to do it by falsely inflating the risks. A falsely negative or dangerous view of the future prevents us from seeing social progress, and thus from generating evidence-based visions and cooperative strategies to create more of that progress. It’s time for us to recognize that we were born into a world, and a universe, that is deeply biased for acceleration to continue. The truth is, we humans don’t seem to be powerful enough to stop this train. We should scale our egos back down to the appropriate size, and pay more attention to the natural systems of intelligence, immunity, and morality that have protected us thus far, and learn how to bring those invaluable systems into our technologies as quickly as we can. That migration will happen inevitably, sooner or later, but sooner would be a lot less painful for all us than later.
Fortunately, since the advent of global decentralized electronic media with the web in the 1990s, the accelerating transparency that comes along with it, the web is rapidly exposing and quantifying our bad actors, and we’re increasingly learning how to shame those actors into reforming their behavior. Add to this the coming effect of both centralized and decentralized AI, as we’ll see in our discussion of the knowledge web and personal sims in this chapter, and one can see that very big counterforces to regressive behaviors are soon going to emerge. Even growing environmental catastrophes as big as climate change will be halted in coming years. We may not have to resort to seeding our oceans with a few hundred million dollars worth of iron, or polluting our atmosphere with temperature-lowering sulfides, as some geoengineers propose, but it’s worth remembering that such solutions exist, waiting to be used if we continue to be particularly greedy and short-sighted. We don’t know why we live in a world where it is so easy to use science and technology to fix the problems that we create for ourselves, but we do.
The futurist Phil Bowermaster likes to say that science and tech have turned all of us 21st century denizens into “wizard kings” armed with vast new digital powers, beyond those of kings and even many fairy tale wizards of old. But as comedian Louis CK famously said in a lovely 4 minute monologue in 2008, “everything’s amazing yet nobody’s happy.” Psychologically, modern humans reserve the right to be unhappy, no matter how advanced our world becomes. That may be human nature, but it isn’t civilization’s nature, which is accelerating. So we have a pretty big blind spot that needs to be fixed, if we want to manage the future well.
The longer we ignore accelerating planetary processes of collective intelligence, immunity, and morality development, the longer our political, economic, technological, and social policies will remain unenlightened, stuck in the past, ineffective, and focused on the wrong goals. The longer we wait to study these processes with the rigor they deserve, the longer we remain burdened with preventable suffering, living in the flatlands below the knee of the next big advance in capacity building, intelligence advancement, and wealth creation.
Even the academic field of Big History, which tells the Big Picture story of complexity emergence, has remained silent, so far, on this Great Race to Inner Space. I’ve had personal conversations with the field’s leaders, and they don’t yet even discuss it as a possible scenario for the future in their introductory texts. That is unacceptable, and failure of nerve on their part. Ultimately, they are going to have to recognize accelerating change, as the Great Race is going to continue, and it will get ever more stunningly obvious as the 21st century unfolds.
A Few Action Items to Start With
What are some action items for leaders with respect to the Great Race? Here are just three to start with, in three broad categories:
- Understanding. We can help individuals, teams, and organizations to better understand acceleration, from both Big Picture and personal perspectives, and realize that we can’t get off this exponential train.
- Foresight. We can help our clients develop a continual learning and evidence-based organizational culture, practice Three Ps foresight, and steer toward universal ideals like the Five Goals, and use historical and comparative perspectives to find better strategies and policies.
- Action. We can help our clients to practice the Eight Skills, to invest more intelligently in the STEEPS factors that create and regulate accelerating change, to commit to higher ethical standards, and to create the most humanizing futures we can.
These are just a few of the more obvious opportunities. Many more action items may come to you as you read this Guide.
As our pace of life continues to speed up, many people and organizations will continue to react with fear and fundamentalism to accelerating change. The 2016 election of Donald Trump as US President is just one of many examples of predictable periodic resurgent fundamentalism, in this case, due to the effects of hyper-speed globalization, under conditions of still-growing plutocracy which have kept America’s middle class stagnant for decades. Such patterns are perennial and easily seen with a historical lens. A similar and far more dramatic and long-lived return fundamentalism occurred in Iran under the Ayatollahs in 1979, also due to too rapid and too unequal socioeconomic change. See my essay God Fights Back: The Return of Religious Fundamentalism to Iran (2012) for some of the sad and sobering details.
In our current times, it helps to remember that Trump, and all our modern presidents, are now just a few of many leading players in the global acceleration game. The more pervasive and powerful our technological systems become, the less important all political roles become, and the more important technical and economic leadership by comparison. Cities, corporations, and technologies now drive our political change, in a more pluralistic world than we had in the 20th century. The biggest issue is our increasingly rapid rates of change, and how we can best collectively steer them toward better ends.
A more mature and responsible media would be focusing on the acceleration story, and would be continually debating ways to make acceleration as broadly empowering as possible, rather than simply relaying the latest news on any of our presidents. The 95/5 rule tells us that the vast majority of change is always being driven bottom up, by each of us. That’s where the greatest stories and action always are. We need to correct our natural bias to overfocus on the 5%, on what our national and our top corporate leaders are doing. That’s easy to do, but it is largely missing the real change. In truth, we are each leading ourselves to better futures, individually and collectively, with the countless decisions and actions we make every day.