The Digital Machine

The Digital Machine

 Melanie Shi

Melanie Shi

What is a machine? Or rather, more aptly, what is the machine?

Hulking, noisy, industrial. Ominous to some, utopian to others. More than anything, though, it is profoundly visible. And with this visibility comes its obvious use as a symbol. As the 19th century Scottish social critic Thomas Carlyle put it, the machine was the “sign of the times.” Carlyle lived through the tail end of the Industrial Revolution and found “the Machine” to be not only the most conspicuous emblem of modernity but, in fact, a microcosm of it. It is hardly surprising that Carlyle took issue with the increasingly mechanized society in which he found himself. Indeed, it is difficult to find writers from any time period who whole-heartedly embraced the advancement of a set of innovations as startlingly new as the industrial era’s were. Slightly more surprising, though, is that to Carlyle, the succession of values had been perverted and reversed: the machine did not inherit the values of the society from which it was born; instead, that society now simulated the values of the machine, which appeared to have come out of some rupture in history rather than its gentle and continuous flow.

Carlyle did not simply worry that workers were being treated more like the machines that would eventually replace them (although this did figure into his criticism). Instead, he had broader concerns about how the ways in which humans relate to each other, to themselves, and to the world around them had changed and would continue to change. Carlyle used the metaphor of the machine to describe how “not the external and physical alone [came to be] … managed by machinery, but the internal and spiritual also.” The physicality of the machine was internalized: the mechanical, the finite, and the exterior assumed a primacy over the invisible, the infinite, and the interior. For instance, according to Carlyle, the highly theoretical philosophy of mind fell out of fashion, while the highly experimental “hard” sciences rose in popularity. And, of course, the rise of the machine predictably led to an increased consideration of efficiency in all daily tasks. The idea of producing something autonomously and by hand became preposterous: Carlyle wrote that even “for the simplest operation, some helps and accompaniments, some cunning abbreviating process [was] in readiness.”

Today, the analogue of the machine must be the computer, and we might think of conducting a similarly Carlylian analysis on this new device. As a symbol for our current times, in what ways is it similar to or different from the machine? It is tempting to claim that the primary difference between the two is simply the type of work that is performed. Machines performed physical labor, and now computers perform higher order operations, what might be considered mental labor – manipulating numbers, interpreting code, storing things in “memory,” etc. However, this explanation does not quite get to the heart of the matter.

For example, imagine a non-electronic device that could perform the exact same operations as a computer, but in a mechanical way. Gears grind, axles turn, wheels spin, and out pops a result. (As an aside, while this idea might initially seem ridiculous, such a device does not strike me as entirely infeasible – see, for instance, the Soviet “water integrator.”) What is interesting about this invention is that it creates for me the impression of a machine more than that of a computer. To classify this imaginary invention as a machine implies, then, that the most salient difference between the machine and the computer does not lie in their differing modes of labor but rather in the fact that the computer performs its work invisibly.* Thus, the demarcation between the machine and the computer lives in the presence or absence of visible internal workings in a device. Take the sewing machine, for example. It acts far more quickly than a human laborer can, but fundamentally, it can be comprehended – it performs the same action as its human analogue (pushing a threaded needle through fabric). The same is true of any system of gears, pulleys, levers, and so on. There is an easily understood mechanical logic available to the user if they simply follow the pattern of movement with their eyes.

A computer, on the other hand, works seemingly by faith and fiat alone. Even if the circuitry were shown to the user (think of the iMac G3 with the translucent back), it would not be clear exactly what is happening when the computer is in use. Unlike the machine, it is inert. This is especially true with respect to the internet. At least with a basic belief in the power of electricity, one can accept that, even if the process is unseen, a computer is taking electrical signals and somehow manipulating them to produce a desired result. There is still a level of locality and physicality to this process: in this understanding, a computer is singular and complete – it performs all requisite operations on its own, in isolation. Once the computer has been connected to the internet, though, mysterious and unknown waves (under the even more mysterious abbreviation “Wi-Fi”) link the computer to others around the world, seemingly by magic.† And we need not even go into “the cloud.”

In addition, the visible/invisible divide has far reaching implications. One particularly social implication has to do with the systems of arrangement with which the machine and the computer are associated. We might think of them as having separate organizational principles. On the one hand, the advent of the machine was closely followed by a corresponding revolution in organization: Ford’s assembly line and the creation of the factory. Suddenly, laborers did not work autonomously but as a collective, each performing a task that would be meaningless on its own. And bound up with the worker’s new role was the capitalist impulse to measure and equate, which resulted in the workers themselves being treated only as units of some universal measure of labor. Ultimately then, this structure mimicked that of its root cause: in the factory, each worker acted as a cog in the machine – regular, quantifiable, and replaceable. In this way, the organization of the machine spread to the very way in which the machine was used in the first place, replicating its logic outwards.

What, then, of the computer? Well, the concomitant organizing principle of the computer must be related to the newly minted concept of “data.” No longer are people the currency in which contemporary capitalism deals – at least not people in the physical sense of the word. Instead, the worker is converted to a cloud of numbers that is subsequently used to identify, characterize, and organize them. These numbers encompass the obviously numerical traits, such as one’s age or weight, as well as traits previously thought to be unmeasurable, such as one’s likes and dislikes. More generally, in fact, any member of society can be quantified in this way and put on file for later use, not just those under the purview of a manager. Thus, following the example of the computer, which seems to operate outside of the physical realm, modern capitalism now attempts to order a population not by organizing the physical embodiments of its people but rather by maneuvering the data that stands in for them.

So it is that the workers of the digital era for the most part no longer stand on an open factory floor, surveyed from above by a member of middle management. Instead, they work in cubicles, neatly arranged in the rows and columns of a database. And, as one might expect, a single number in this vast database matters little to the end result, the ultimate calculation. Far from comprising independent yet interrelated gears that help the larger machine turn, – gears that, while replaceable, could nonetheless still bring the operation of the machine as a whole to a grinding halt – the operative system is now an expansive and invisible network, encompassing not workers but internet users. The function of a single node in this network is only as a representative of how other nodes will act; each serves as a metonym for some subset(s) of the network as a whole. Thus, alone, an individual is useless (a single node going offline does not a catastrophe make), but as an indicator of broader trends, they are invaluable. Facebook does not care about what you are interested in – it cares about what people like you are interested in.

Indeed, we can never truly know what role we play in the larger order of things. The factory worker understands (and can see) that their small task of measuring out wire will allow for that wire to be straightened, cut, pointed, ground, and eventually turned into a pin: all discrete, visible operations. Moreover, they are considered, for all intents and purposes, the same as every other worker, each drawn from some standing reserve of identical laborers. Today’s internet user, on the other hand, has no conception of how they are being interpreted by and assimilated into the larger network. They often do not even know that they are being used, that their data is being collected at all. Further, in their new role, they are encouraged to act as differently from each other as possible. Individuality is emphasized so that unique signs can be cached out for new markets, so that deeper and subtler classifications can be made in order for new demographics to be explored and mined. I am no longer just “male, age 18-24.” Now, I am “male, age 18-24, in college, living in a city, interested in philosophy,” and the list goes on and on. I can likely be placed into any number of subsets. Considered altogether, this system is oceanic, dynamic, and unknowable, and, once again, the computer’s own innate structures (as a function of its invisibility) are multiplied and magnified outwards to the society it inhabits.

It should be noted that this new technological enframing is not necessarily worse than the old, just different. There are clear advantages to being a node instead of a gear. For instance, regardless of the system that one’s fostered individuality might inadvertently support, it is undeniable that the node has more freedom than the gear. One is no longer interchangeable with others, and there are real social (not to mention psychological) benefits to being able to choose your own way; the archetypal worker of today, for instance, is not forced to wear a uniform.

My issue, however, is that our understanding of these structures has not caught up with the reality. We continue to use the language of the machine to describe what is now definitively the world of the computer. We still call bureaucracy a machine and the worker within it a cog. We still speak of putting one’s body upon the gears and, (in the mid ‘90s at least) raging against the machine. Even that most treasured dictate of the proletarian revolution to “seize the means of production” relies on a conception of capitalism’s apparatuses as fundamentally physical and visible. We might ask ourselves, what are the digital “means of production,” and how can data that is gathered in unfathomable quantities and processed at unfathomable speeds possibly be “seized”?

This mislabeling leads to a more general worry as well. Without the ability to effectively describe the frameworks that make up and order much of our daily lives, we lose some of our capacity to interact authentically with the world around us. We spend hours each day online, participating in this grand network, yet we relegate the computer to the realm of experts and pretend it is of no concern to us how it works. Emerson, discussing the danger of metaphors that no longer have a basis in reality, writes in his essay “Nature” that when “new imagery ceases to be created, and old words are perverted to stand for things which are not … In due time the fraud is manifest, and words lose all power to stimulate the understanding or the affections.” He calls these perverted images “paper currency” for when “there is no bullion left in the vaults.” Perhaps today they are more aptly compared to Bitcoin.


* For a discussion on how the invisibility of this work might, in fact, be related to our conception of the computer as performing “intelligent” tasks and, more generally, on how there exists a social dimension to “intelligence” that is associated with the concealment of work, see Alexander Garnick’s excellent piece “Sleight of Mind” in this same issue of the Gadfly.

† It might be argued that these “waves” are really no different from the radio waves of old. However, I would argue that the radio has a physical presence that the internet does not share. Radio towers loom over cities and implicitly provide reassurance that the radio does indeed have a physical backing to it. Data centers, on the other hand, are hidden away in low population areas, on the periphery of the public’s gaze. Moreover, the radio seems to be, on the whole, better understood and more accepted than its younger kin. Many understand the basics of transmitters and receivers; few truly understand the layers of the internet that allow a two-way sharing of information to occur.

Interview With Meghan O'Gieblyn

Interview With Meghan O'Gieblyn

Notes on Automation

Notes on Automation