Digital Thinking
The construction of abstract binary representations is equally noticeable in human and digital machine intelligence. We suggest that this constructive mode, or the dependent arising of binaries, carries implications for both the scope of simulation and the status of the transcendent.
Radical Simplification
Digital computation is performed based on binary code—absence and presence, zero and one. Algorithms act on combinations of such data to create models, representations of what would be the case in the world under the circumstances stipulated by the algorithm. This setup lends itself to an account of, for example, human cognition. It is compelling to think of our minds as receiving stimulating input from the world, input that once received is processed by a controlling agency, thereby producing outputs that can be integrated into our ongoing cognitive construction of the world. It is also obvious how such an account of mind can seem impoverished, if not self defeating, for it may seem more an account of mindless processing than of genuine cognitive engagement, as with consciousness. Yet going along with the idea of computers and minds as Turing machines, we will here employ a Buddhist analysis to further explore the potentials for alignment and integration between intelligent organisms and machines.
As we just noted, the construction of a computational model rests on binary signals that are interpreted as digital data, 1s and 0s. At first glance, it may seem outrageous to suggest that consciousness could be brought down to that level of crude simplicity. How should it be possible to reduce, as it were, the entirety of human civilization to elaborate constructions built on rudimentary signals along the lines of simple presence and absence? Even if we insisted on doing so, would not the in-our-face richness and sophistication of lived experience still remain to belie our efforts? Unless the idea of mind as a digital computer system were proposed in some weak allegorical sense, and within a very specific context, would not the seemingly irreducible richness of experience render such a theory absurd?
At the same time, perhaps something similar could be argued just as well with respect to the physical processes that are at work in a computing machine. Claiming that an account of the simple presence or absence of electrons meaningfully captures the totality of all subatomic particles and processes in the machine seems quite outrageous. Even if we would insist on going through with such a stark reduction, would not a simple nod toward the rich complexity of the quantum world, for example, be enough to expose such an effort as hopelessly naive? But to in this way criticize the very idea of digital computation as wholly unrealistic would of course be absurd. The world quite obviously abounds with machines that, for most intents and purposes, must be described as “digital computers.” Moreover, those machines increasingly contribute to the way our world is shaped and experienced. A simple acknowledgement to be made here is then that of the power of radical simplification.
Toward the Interdependent Foundations of Code
More specifically, however, it may also be possible to notice, both analytically and phenomenologically, a form of binary code that shapes the acts of thinking that characterize human minds. With the help of the philosophy of dependent arising that we began to consider in a previous post, the features of that structuring can be made sharper and more noticeable. To begin such an analysis, let us first return to the 1s and 0s that make up digital data, noticing how they stand for full presence versus total absence. There are no tones of combination or compromise possible here: The signal is either there or not there. In this way, the unfractured totality of 1 contrasts with the utter nothingness of 0. 1 and 0 are as different from each other as anything can possibly be, and yet when 1 or 0 is repeated, the repetitions are all completely identical. Now, such a pair becomes possible only by mutual contrasting: Neither 1 nor 0 can be found in isolation, because their identification depends wholly on the availability of the contrary concept.
In Buddhist philosophy, this point is emphasized in, for example, the Analysis of Fire and Fuel, the tenth chapter in Nagarjuna’s classic Insight: Root of Middle Way. Pure difference and sameness cannot be encountered anywhere—except in terms of such mutually dependent imputation. No concrete act of differentiation or identification can involve elements that are wholly different or identical—except as pure abstractions. In order for two things to be deemed different from each other, they must, at a minimum, be the same with respect to being accessible for comparison. Likewise, any act of identification requires an element of differentiation that can subsequently be claimed cancelled through equation. Even as abstractions, “same” and “different” require each other to work.
Binaries such as 1 and 0 are conceptual constructs of exactly this kind. Despite their apparent opposition, they cannot be considered truly different because each requires the other to be what it is. At the same time, precisely because they are binaries, these concepts become intelligible only if we grant them powers of complete mutual exclusion. Such are then the foundations of digital code: 1s and 0s that are really neither the same nor different.
Toward a Transparency of the Transcendent
In our previous post, we noted that the teaching of dependent arising implies that nothing is what it is independently. All whatever there may be requires the availability of things other than themselves to stand out and appear the way they do. Or in order simply to be. The case of “same” and “different,” or 0 and 1, illustrate this point well. But what then to think of the things that are the concrete foundations of such concepts or codes? Take for example the electrical signal that may be either there or not. We can indeed speak and think of that grounding signal with such sophistication and success that our doing so enables the running of an intelligent machine. Is this not evidence that our concepts—be they as they may under philosophical analysis—somehow manage to connect with deep structures of reality? Well, it goes without saying that the signal for encoding requires factors other than itself to arise, and those factors are again not independent. They all need a range of other factors to come into being and be identified the way they are. The fabric, as it were, of dependent arising extends in this way infinitely. More specifically, all the relevant elements that we may determine along the way, from start to goal, become recognizable exclusively through the use of concepts such as same and different, presence and absence. The framework, as it were, of dependent imputation thus extends pervasively as well.
On the question of the deep structures of reality, these states of affairs do of course not render reality any less profound. But the profundity of the real world can then not meaningfully be located in a realm of transcendence. The real world is not beyond us, because it emerges with us. The implications for computer modelling appear profound.
Although 0 and 1 may seem odd candidates for ontological rock bottom, that can hardly be the case from a perspective of dependent arising. Human minds develop and employ their intelligence through the application of binary concepts of sameness and difference, presence and absence, as do digital computers. This is how we collectively develop our models of the world—a world that provides input for our constructions and hence cannot be seen as intrinsically separate from them.
Let us finally note that if the game of ontology ends with dependent arising, then this would not permit any reification of the principle itself. In terms of computing, although infinite models can be developed based on binary code, the algorithmic code is only intelligible as such in context and in dependence on things other than itself. While the principle of simulation may capture well the constructive processes that characterize both human and machine intelligence, the program of digital physics has no purchase as an ultimate metaphysics. The real word is readily available—in context.
