Review of My Book “The Software of the Future”

Florian Arnold

The review originally appeared in the Vilem Flusser Archive at the University of the Arts, Berlin.

Human culture has always grappled with an “other” aspect of itself. Until recently, it was still generally accepted that this “other” was “nature”—understood as the pre-cultural origin of culture and, perhaps, its conciliatory end. Industrialized modernity, in particular, kept nature untouched by humans in reserve—it even created nature reserves—and nurtured it as its own guilty conscience. Since the middle of the last century, however, a development has been emerging whose contours are now clearly discernible. Another “Other” is looming on the horizon of human culture: artificial life or artificial intelligence.

With this panorama in mind, there have recently been increasing attempts to discuss, promote, and critique the implications of this technology-driven development for the human world. Discussions about AI, big data, or the omnipotence of algorithms permeate the arts and culture sections, and political programs such as “Industry 4.0” are instilling a spirit of optimism even in otherwise conservative government circles, whereas literary bestsellers like Dave Eggers’ *The Circle* once again cast suspicion on California’s technological optimism as mere ideology. All these debates revolve—as is only natural at first glance—around humanity, its hopes, and its fears. But what would become of this anthropocentrism when it comes to a virtual future, if one of its decisive consequences were precisely the actual emergence of a new species—a truly different “Other” beyond or apart from the human?

With *The Software of the Future*, Alan Shapiro presents a pamphlet that places this question at the center of the discussion for the first time. The thoughts in this text no longer revolve solely around the human future, but equally around the independent existence of things that may yet come to be. Arguing against subject-centrism in culture, technology, and the future, Shapiro advocates for a reorientation of perception: It is about object orientation, object-centeredness, and ultimately an “empowerment of objects,” a “realization of artificial life” (p. 8). True to the subtitle, “The Model Precedes Reality,” Shapiro draws on Baudrillard to proclaim a new paradigm within computer science that will ultimately leave nothing untouched.

The evolution of programming languages is vividly illustrated: Starting from procedural languages and moving through object-oriented ones, we seem to have already traveled halfway toward an ultimately object-centered programming language, in which the software ultimately becomes the programmer of itself. We are on the verge of enabling artificial life; we are, as it were, programming informational autographs—yet to this day we lack a satisfactory theory for learning to understand artificial life as a distinct form of life. Much like our relationship to the natural world of animals, we still seem unwilling to relinquish our claim to be the dominant *animal rationale*. How much more difficult, then, must it seem to even acknowledge artificial life—of which we simultaneously consider ourselves the creators—in its own right?

According to Shapiro, to find answers to these questions, it is no longer sufficient to pursue computer science solely from a natural- or engineering-science perspective. Instead, Shapiro advocates for radical transdisciplinarity. One of the crucial shortcomings of computer science—not only in a critical sense but equally in a productive one—is a persistent neglect of the humanities and social sciences that continues to this day. Only by combining the strengths of all disciplines relevant to life can we succeed in completely breaking free from an outdated binary-mechanistic paradigm within AI research and open up new—and, above all, actionable—perspectives. A shift in thinking—not least a philosophical one—seems to be the necessary prerequisite for this.

Shapiro himself demonstrates how this might happen: If we are ever to succeed in programming an AI that is at least on par with humans, we cannot rely solely on the current imitative methods of production. Programming should not aim to reproduce the “programmer” (such as the human brain). Instead, the goal lies in object-centered creativity. (p. 20). And this step forward, in turn, requires a step backward: Only when we apply programming to artificial life does the prospect of an autonomous AI even become a possibility. Artificial life, however, seems possible only where contingency and indeterminacy are already incorporated into the programming process:

What we need is a new technique that reinforces the inheritance mechanism of object-orientation. In addition to inheritance, we will establish analogies and similarities between the software instance and its blueprints—the software classes—which endow the instance with its capabilities. The most fundamental analogy defining the basis of this subsystem’s operations is the idea of a software instance that has a choice—that possesses existential freedom and is not determined by existing attributes and data. The possibility of choice is made possible by incompleteness. Incompleteness, in turn, is made possible by the possibility of choice. […] Whereas Descartes and Leibniz sought to derive a system of knowledge from a few established certainties in the 17th century, we will begin with incompleteness. (29f.)

Artificial—but also natural—life cannot simply be constructed because it continually deconstructs itself; that is, it proceeds in a “deconstructive” manner, it constantly reinvents itself. Thus, French poststructuralism serves as Shapiro’s inspiration for his project to decentralize the Cartesian subject-centrism of computer science and to recentralize it within the software object itself. Thus, before the human programmer can nominally become the author and creator of artificial life, he must already have seen through his own authorial fiction of a “programmer-god.” The point here is to understand one’s own programming activity, in turn, as a program that itself already seems programmed to program artificial life. Ultimately, this radically calls into question the notion of programming as a form of prescribing or dictating. No one programs anyone else without simultaneously being programmed themselves. Programming does not mean exercising dominion over the Other, but rather enabling the Other’s freedom—even if this Other and its freedom would ultimately no longer be the Other and the freedom of the human being.

Shapiro seems committed to helping this essential insight gain traction—both theoretically and practically—as he presents his far-reaching diagnoses and prognoses, at times in an promotional tone, at other times in a critical one. If one were to define a genre for this text, it could most likely be described as a “manifesto.” It already seems predetermined here what lies ahead for us, should we not wish to rid ourselves of the whole matter once and for all. At least for now, everything points against it—apart from the fact that we ourselves no longer seem to be the only ones with a say in the matter.

One must give Shapiro credit for managing, with just a few strokes, to paint a future before our eyes that goes even beyond our present hopes and fears. Yes, in all likelihood, another species will indeed evolve, and yes, we ourselves will be responsible for it. But the greater hope lies in the fact that we will lose our fear of precisely this—and it must be emphasized: our fear. Artificial life will ultimately be artificial life, just as human life is human. This seems to both surpass and fall short of Gilbert Simondon’s or Bruno Latour’s demands for political equality:

“In a certain sense, Simondon committed a fallacy that was as anthropocentric as it was anthropomorphic when he attempted to defend the honor and pride of technical objects, since he attributed a kind of human status to them. But this very thing would negate their ‘otherness’ from humans […]. It would be more accurate to state that technical objects are alive, rather than assuming that only human life is worth living. Or better yet: that they are granted a status that transcends the binary opposition of alive or not alive. The focus should be on humans and technical objects coexisting in their diversity—their ‘otherness.’” (p. 60)

Shapiro’s reflections make it clear that we would in fact be dealing with a different Other—an Other that would no longer serve as a projection screen for our own fears and hopes. But perhaps this is precisely where our true, our greatest fear lies: an unmediated encounter with our fear of ourselves—as our own Other. It is to be hoped that we will one day lose this fear along with our fear of the other Other.