[This article is available as a preview of my upcoming book on the subject of human and non-human machine consciousness and its impact on our future society. Please join my newsletter email list to be notified when the book will be available. Your comments are always desired!]
Human kind typically lacks a functional understanding of the long-term consequences of our actions. The corollary to this statement is that we also do not comprehend exponential growth. The math is learnable, but the brain is inexperienced in estimating mathematically exponential effects in our environment, such as the growth rate of virus exposure during a pandemic. Although the appearance of exponential growth rates in the physical world is a relatively recent phenomenon, the mathematical concept has existed in the form of logarithms, first discovered in the early 1600s. Astronomers of the time successfully used logarithms to perform calculation shortcuts (does anyone remember using slide rules?).
Today, by underestimating the impact of exponential growth on our world and making the corresponding mistakes, do we resemble a more primitive being? Is this desirable? Humans are still able to estimate or predict many events accurately; how quickly people learn, how fast organizations change, the growth rates for institutions, etc. as these systems do not typically exhibit exponential growth. However, to an ordinary man, the proliferation of exponential growth patterns represents a perceptive mismatch that easily hampers understanding of exponential growth rates and their place in the 3D world. It is not unfounded to consider this issue at the core of many problems in modern society, especially in understanding how rapid changes in the role of technology and information have played in producing fear in the modern world.
Since World War II, the exponential growth rates of technology and information have been mostly out-of-step with man’s ability to comprehend such change, resulting in output discrepancies. In many aspects of our productivity, the output gap has been progressively filled by the role of computation, computers, and Artificial Intelligence (AI). Simply stated, people cannot physically or mentally process large amounts of information and data fast enough to quickly build accurate models of reality that are needed to make predictions and therefore profitable decisions. Responsibility for that task has been shifted to advanced computing technology and AI. Just keep in mind, data collection is extremely valuable as it’s needed to build models of reality in software simulations that are designed to train AI to respond effectively to real world signals. Without data, AI would be useless.
By the 1970s, Chaos Theory was validating this fact by clearly indicating that successful modeling of reality to make predictions requires extensive data sets and intense computation, supplemented by graphical visualization of the results to aid comprehension. The application of Chaos Theory and other mathematical principles in making decisions has manifested great monetary benefit in the financial world. The monster investment bank of Goldman Sachs is well known internationally for its application of computer simulations in analyzing market and investment conditions, producing fantastically profitable results. A rapid expansion in reliance on computers to do much of the decision-making work for us, identifying and analyzing the presence of exponential functions in our reality, has made many jobs obsolete and humans routinely declared superfluous to machines. Unfortunately, the success of computer predictions has also effectively marginalized our intuitive abilities. Failing to acknowledge and act on the effects of exponential growth has contributed to the contemporary atrophy of the human ability for making decisions.
Now, let’s not forget that in the real world an exponential curve does not go on forever. What it transforms into can resemble something called an “S” curve with its associated inflections, or changes in direction. Computers and AI are exactly poised to aid us in interpreting the data of exponential systems. Take for example the paradigm of computing power measured by processor speed. By many credible reports, growth in computer processor speed has been exponential for decades. Consider that at the top of a tech corporation, the decision-making food chain mandates that economics be paramount; growth in processor manufacture is limited by a projection of its profitability. As an example of the use of technology to make decisions, there are also many other factors besides processor speed (and profitability) to consider. In addition to a multitude of market factors impacting profitability, for the analyzing the use of processors in computers systems we could include 1) processor speed, 2) the energy usage (power demands of processors are decreasing), 3) employing parallel computing (using multiple processors in a single system is increasing), 4) changes in the types of processors (increasing use of non-floating point GPUs), and 5) increased efficiency of processor manufacture (improvement in hardware design and production techniques). So, it’s apparent that there is limited capability to examine, without the use of computers and AI, the impact of all these parameters on a decision, especially since many of the factors have exponential growth rates and unique “S” curves or inflection points. It’s easy to predict that the use of AI is only going to expand in the technology industry.
For human related functions, some argue the ultimate bottleneck to AI growth will be limitations on the amount of human experience that can be collected in existing data. Without changing the source or form of the data collected, it would seem possible at some point to eventually amass all data collectable on human experience. But, today we are in the middle of an exponential growth rate in data collection from individuals, so that point in time is still a ways off. However, today this data is being used globally to train predictive models of reality for deployment in AI systems. If there is an upper limit on data collection, where does this lead us? Knowing that there is exponential growth in data collection, it seems plausible that in the future we’ll get into an inflection point where a limit is reached in the amount of data collectable, for a given source and form. This would effectively put us into “S” curve territory and up-end the exponential growth rate in human derived data used to model reality for AI.
By changing the type, source, and form of data collected, analysis of this data for reality model building for AI would potentially move technology into a new paradigm of creation. This transition could be viewed as a point of departure that opens up into new arenas the resulting technology realized from modeling of data collected. Motivated by not only changes in data collection methods, technology would be transitioning beyond incorporating only human life experience and into a realm that considers all life and reality. We could actually reach a stage where the models of reality that AI are based on are created from data derived from analysis of experiences and events that are outside a framework of human experience, perhaps delving into non-human reality (such as off-world existence, an entirely different subject of exploration). It’s implied here that we must ask what form of spiritual awareness would be integrated within the model of reality embedded in the resulting AI, or would it be considered Artificial Consciousness (AC). In this future scenario, we could be looking at androids operating with an evolved awareness or consciousness, a broader perspective on the nature of all life and reality, and perhaps more accurately representing humans for what they really are; spiritual beings.
In this future, we are brought decisively to a critical inflection point, likely necessitating legal action and the creation of new laws for the technology industry. The objective would be to protect and preserve our human identity when it is represented in the behavior of technology systems, especially when any AI/AC will exist independently or autonomously from humans. By spawning or birthing a new form of physical humanoid machine, human society would be tasked with the mandate to accurately represent the consciousness of a human in machine form. All AI/AC creations using this advance technology (including what is today called General AI, technology closely mimics human behavior) could be required to incorporate a moral/ethical compass that includes a spiritual awareness component. The letter of the law and resulting behavior of such a machine would hopefully both reflect the best qualities of a human being. Arguably, this would be the most significant moment in human history; official legal recognition of Artificial Intelligence (AI), Artificial Consciousness (AC), or Artificial Life (AL) as human technology creations that must represent the best characteristics of a human being. What does all this mean for us today?
Under the best possible conditions, the inherent limitations of information and technology needed to produce AI that is modeled accurately to a human being, may ultimately force the trajectory of AI development away from the predominately profitable motives of technology corporations and back onto a track that is beneficial to human society. In this future scenario, AI/AC/AL could only exist within a legally defined context that requires it to faithfully represent the best of a human. Will the future moguls of tech corps be committed to making this happen? At present, we could only imagine this being partially possible. As a precursor to legal definitions, perhaps the first objective would be to concisely identify what is spirituality, human life (or a human being), our mission on Earth, and the true nature of all reality? Ultimately, answering these questions will be necessary for the expansion of our consciousness and the eventual integration of humans into the universe of life (preceded by our technology). That’s a whopper of a task list!
Lastly, my non-physical exploration of the condition where AI/AC/AL exists successfully in future society says, “Don’t hold your breath”. Profound alterations in the trajectory of humanity can be sudden, often accompanied by violent actions, or they can take a while. My feeling is perhaps a hundred years or more must pass before a major shift is accomplished in these objectives. Needless to say, there are many other factors at play, and the path ahead is paved with uncertainty.
One reply on “Exponential Growth, Technology and the Future”
[…] In 1956, the Dartmouth college summer academic workshop was held to openly clarify concepts of “thinking machines”. The organizer of the workshop, John McCarthy, coined the term “Artificial Intelligence” to describe this new field of investigation. Then, in the early 1960s, research into mathematical modeling was being conducted using computers that eventually lead to the development of “Chaos Theory”. The upshot of the collective academic work around this subject was to firmly establish a new branch of mathematics intimately linked to computation. Chaos theory identified the exponential difficulty of accurately modeling or forecasting complex or chaotic systems of nature (such as weather). It established that successful forecasting required enormous amounts of data and extensive processing power to simulate effectively. Concurrently, the role of fractal patterns and the diverse occurrence of fractal geometry in nature became apparent (please explore influence of the Mandelbrot set), paving the way for more accurate modeling of physical structures in our world. These examples, Chaos theory and fractals, have led to major advancements in the modern implementation of AI modeling which are accelerating the growing influence of technology. The resulting manipulation of limited human perception is no small matter to consider as a consequence. Understanding the exponential growth needed to continue this trend is discussed in my previous article on Exponential Growth. […]