Biocentric to Technocentric Trajectories
The View from Oregon — 273: Friday 26 January 2024
In last week’s newsletter I discussed some of the barriers that machines face in converging upon human-level intelligence and agency, and, if they do begin to converge upon human-machine equivalence, the cognitive uncanny valley they must pass through, or, in passing through this cognitive uncanny valley, they may enter into a trajectory of development that ends human-machine convergence, which then becomes a radical divergence. Here we have two scenarios for the growth and development of computer technology: 1) convergence upon human-machine equivalence, and 2) initial convergence followed by subsequence divergence. What do these two scenarios suggest for the development of a technocentric civilization through the replacement of biological constituents by technological (artificial) constituents?
Between a purely biocentric civilization — in which materials are natural and concepts are qualitative — and a purely technocentric civilization — in which materials are artificial and concepts are quantitative — there must of necessity intervene a transitional period. Arguably, we are already in this transitional period, and we entered this transitional period as soon as we began manufacturing tools, bringing the technosphere into being. However, the replacement of qualitative concepts with quantitative concepts is much more recent. We can identify intimations of the quantification of our conceptual framework in classical antiquity, but the process does not begin in earnest until the scientific revolution. The transition of the conceptual framework, then, already exhibits at least two thresholds, one in classical antiquity and an inflection point after the scientific revolution, and we could plausibly argue for another inflection point starting in the middle of the twentieth century with the advent of computing technology, which forces developments in conceptual quantification through economic incentives to increasingly automate the industrial infrastructure.
Trying to think this through initially from a purely formal perspective, we can identify a number of formally distinct dimensions of developmental trajectories, each of which could be quantified:
- The beginning of the transitional period, however we define this point of origin.
- Developmental thresholds during the transitional period, which define subordinate periods within the larger transitional period.
- The length in time of the transitional period, including the length in time between thresholds embedded within the transitional period.
- The completion of the transitional period when replacement is exhaustively realized.
- The structure of the transitional period, including the structure of subordinate periods within the larger transitional period, each of which might have a distinctive structure. By “structure” I mean some quantifiable property of the replacement process, such as linear, monotonic (non-decreasing), non-monotonic, or punctuated development.
Some of these already quantifiable properties of replacement (understanding that replacement corresponds to the emergence of a technocentric civilization), can be further broken down into distinct scenarios. Most obviously, the length of time of the transitional period (no. 3 above), can be divided into several distinct scenarios, with the divisions between them being conventional. The convention we establish simply needs to be a serviceable way to think about the transitional period. Here is an obvious way to go about this, utilizing short, medium, and long transitional periods:
- Short — 01 to 1,000 years (human time scale to civilizational time scale)
- Medium — 1,000 to 1,000,000 years (climatological time scale to geological time scale)
- Long — 1,000,000 to 1,000,000,000 (one million to one billion years)
Again, the schema we adopt is a convention, and it might be more convenient (or conceptually necessary, depending upon what we are attempting to accomplish) to adopt a different or a more refined schematization. For example, we might choose to make a five-fold schema that adds a shorter and a longer time scale:
- Very short — 0–1 years (nearly instantaneous in historical terms to a human time scale)
- Short — 01 to 1,000 years (one year to one thousand years, or a human time scale to a civilizational time scale)
- Medium — 1,000 to 1,000,000 years (one thousand to one million years, or a climatological time scale to a geological time scale)
- Long — 1,000,000 to 1,000,000,000 (one million to one billion years, or a geological time scale to a cosmological time scale)
- Very long — 1,000,000,000 to 1,000,000,000,000 (one billion years to one trillion years, or cosmological time scale to eschatological time scale)
The more refined our time scale divisions, the more fine-grained our analysis, but the more unwieldy the resulting analysis, and our likelihood of losing the attention of anyone who is looking for an enlightening way of thinking about the problems that confront us. We have to limit the number of categories we set up, but, at the same time, we need to accommodate the most important scenarios that need to be placed within this schematism, and it would be best to place distinctively different scenarios in different boxes. For example, in “hard landing” singularity scenarios, the transitional period is short, or very short. We can add the “very short” category to accommodate this possibility, or we can throw the “hard landing” scenario into the same box as those scenarios that unfold over hundreds of years. If we limit our categories, we conflate scenarios; if we allow more categories, the analysis becomes overly complex. I run into this problem with civilization all the time. Any inherently complex phenomenon like history or civilization or life lures us into this dilemma. One way to address the dilemma is to adopt a couple of different frameworks, one simpler, another more complex, and use whichever is needed in a given context to make the point.
The point of the above analysis is that, once we break down the problem of the transition from biocentric to technocentric civilization into discrete chunks and structurally distinct scenarios, we (hopefully) have a handle on the various possibilities, and some of these already fit what we know so far about our historical development. Above I noted an initial period of technocentric transformation which begins with the invention of tools (about two million years ago) up to the scientific revolution, so this is a process that has unfolded over about two million years and thus already fits into the medium or long temporal scales. However, if we limit ourselves to civilizations, then we are only talking about an initial transitional period from the advent of civilization (about ten thousand years ago, give or take), which means the first stage of the transition from biocentric to technocentric civilization lies on the medium time scale, but it also means that the technological replacement process begins long before the advent of civilization.
The next period from the scientific revolution to the advent of computing technology is about four hundred years (from the publication of Copernicus’ De revolutionibus orbium coelestium in the middle of the sixteenth century to the introduction of electromechanical computers in the middle of the twentieth century), putting it in the short time scale. That is a considerable difference between the first two subdivisions within the transitional period from a biocentric regime to a technocentric regime. The next subdivision, from the advent of computers to whatever threshold will present itself in the future, might occur on any one of the defined time scales; we simply don’t know how long it takes for an unprecedented development to unfold. There is no historical parallel to the development of computer technology (unless we water down the idea of an historical parallel until computers resemble some past technology).
The beginning, end, duration, and subdivision of the process of replacement are all problems of periodization; the structure of the transitional period introduces its own complexities, and these will be predicated upon the analysis of technology that we adopt. In the philosophy of technology there is no agreement on a definition of technology, and that means that there is also no agreement upon a taxonomy of technologies, which is what we need in analyzing technological development in this kind of abstract way. It is, again, much the same in the discussion of civilization. Everyone has their own preferred definition of civilization, and no two definitions are exactly alike. There is no consensus definition that has allowed a Lakatosian scientific research program to emerge, and thus for the study of civilization to reach the paradigmatic stage. The same is true of the study of technology; the most basic distinctions are missing, and in order to work through the possible permutations, even at the most abstract level, we have to construct the non-existing conceptual framework necessary for the task.
For example, if we take technological development within a given periodization, are we going to treat it as a whole, or divide our analysis into classes of technology? Some classes of technology that are especially relevant to civilization, such as farming and weapons of warfare, exhibit a monotonic growth in sophistication over time, even given social disruption and dark ages. Plowing technology continued its slow, incremental development throughout the dark ages, since civilization, once clustered around the Mediterranean, expanded northward into the wet, clay-heavy soils of northern Europe, it was forced to adapt and innovate continuously; this expansion would not have occurred if Roman farming methods from North Africa had been iterated unchanged in different biomes.
During this same period, however, urban centers lost the ability to manufacture machines like the Antikythera mechanism, so there were technological losses during the same period. If we look only at plowing, technological development is monotonic; if we look only at refined astronomical instruments, technological development is non-linear throughout this period. If we take technology on the whole during this period, we must include the losses, and development would then be presented as non-linear. This simplification conflates technological taxa, and can give a false view of technological development. If we analyze technology into taxa, the analysis becomes more complex, but also more accurate. How we analyze technology into taxa will also bear upon the analysis.
Given an analysis of periodization, and a taxonomy of technology, we can produce an exhaustive analysis of the permutations of the possibilities for the transition from biocentric to technocentric civilization. If we have adequately accounted for all relevant factors (and we certainly may miss something important, as we usually do), then the future will exemplify one of these permutations. If we can also formulate appropriate and realistic responses to and interventions for each permutation, once we can recognize what scenario is unfolding, we can attempt whatever interventions we have determined are appropriate. That may give us the edge we need for a potential great filter as we converge on human-machine equivalence. No scenario comes with any guarantees, but there are at least analytical possibilities to constrain and maybe tame the problem.
