In his controversial Love and Sex with Robots (2007), international Chess Master and renowned Artificial Intelligence expert David Levy strikes timely and telling cultural and technological parallels in this declaration:
When robot creatures are generally perceived as being similar to biological creatures, the effect on society will be enormous. It will be as though hordes of people from a hitherto-unknown and far-off land have emigrated to our shores, a people who behave like us in many ways but who are very clearly different.1
That he establishes these parallels so easily has been less controversial than his book’s assumptions of and possible impact on gender relationships, and his nonchalant relationship to ethics. However, the ease of these parallels and how natural they are to even the most casual reader is worth examining. As argued here, it is through these facets that we can make sense of how and why the very question of ethics has become central to public conversations about human relationships with machines.
Racing the Singularity
The history of technology, from industrialization to cybernetics, from robotics to AI, has in fact been replete with this type of casual parallel between humans and machines, non-whites and robots, technology and social, cultural “others.” The history of Science Fiction has served to rationalize and naturalize those parallels in such ways that issues of race and difference—including sex and immigration, as Levy reminds us—are central to the genre’s software, so much so that they always already make sense in our consumption of such narratives. It is not important for the author to grasp how remarkable it is that the primary template for understanding robots is that of race, sex, labor, and immigration, because this has always been necessary. The real-world impact of these metaphors has become so naturalized as to be culturally invisible. This, after all, is the logic of metaphor: in making connections that already make sense, metaphor erases the historical tensions that make them possible and continue to give them charge.
Rastus Robot, “the mechanical negro”, developed by S.M. Kinter (pictured here re-enacting Swiss folk hero William Tell), 1931.
Most important here is the fact that Levy narrates through sex, race, and robotics what we increasingly call “the singularity,” after futurist/technologist Ray Kurzweil: that apocalyptic moment where machines equal or transcend humans in their cognitive abilities and, perhaps, social value; that moment that has fueled science fiction from its origins, in the work of writers such as Samuel Butler and Karel Capek. These authors, with their “revolt of the robots”/”rise of the machines” trope in the wake of Mary Shelley’s Frankenstein; or the Modern Prometheus (1818) and a lesser-known story like Herman Melville’s “The Bell Tower” in The Piazza Tales (1856), shape the genre insofar as it grew directly out of 19th-century anxieties about slave reprisals, fears of colonial resistance, and 19th- and 20th-century terror of labor insurrections.
In our moment, however, the singularity is defined and produced largely by what Black British theorist/artist and notable Afrofuturist Kodwo Eshun calls the futures industry—“the intersecting industries of technoscience fictional media, technological projection and market prediction.”2 Taking Eshun’s perspective in which race is central to how we imagine the future, what if we do not depend on this industry and its distorted sense of the future as inevitable and instead draw from that growing cluster of perspectives (I hesitate to yet call it a discourse) called Afrofuturism to make sense of the singularity? Rooted in the historical experiences of black peoples in and after slavery and colonialism, there is an eschatology there that removes us from that turgid binary of a corporate future and an allegedly transcended past. It suggests this: What if the apocalypse has already happened? What if, as musician and self-styled prophet Sun Ra would have it, this is already after the end of the world? What if the anxiety that many have about the future—with sentient machines and an Artificial Intelligence that dwarfs human conceptions of God, state, or infinity, and that also promises pleasures and freedoms dwarfing any we have previously known—is merely déjà vu?
There is something liberating in this take on time. It reorients our thinking away from a fear of the imminent to an acknowledgement of immanence, of being already entangled with the repercussions of that fear. It removes us from the rigidity of state sponsored techno-utopianisms and situates temporality in the more familiar though deeply unsettling realm of ongoing cultural anxieties and longings. This reveals that in the context of technology and its corporate futures, there is no Kurzweilian singularity on the horizon, no decisive moment when “humans” transcend their “biology” and technology outstrips human intelligence or when machines surpass human capacity. Or rather, that there is nothing singular about that singularity, hence the fact that black responses to apocalypse are often characterized by a familiarity that verges on blasé. This is because for us the world has already ended. To paraphrase Samuel Delany, the great Black Science fiction writer/critic and theorist of unspeakable, border-crossing desire, apocalypse has come and gone and we are just grubbing in the ashes.3
This eschatology matters because Levy’s future, like Kurzweil’s or Elon Musk’s or those still driving much of the technological development in the West, is merely an echo of an earlier but no less momentous transformation that we have lived with for long enough to forget it had an origin. It is a transformation we recall differently depending on our placement on the human/inhuman divide. From differential positions, that first singularity allowed either a gaining or a losing of power once the boundary between the two categories began to smudge and once the social structures that enabled them began to bear the brunt of challenges to their legitimacy.
With Afrofuturism in mind, then, the imminent moment when machines become citizen, separate and/or equal, or when things become people and emigrate to our shores, will merely reiterate a not so distant past. This is one where blacks, reduced to object status, denied souls and intelligence, poised by the logic of slavery between beast and automaton, between animated tools (as Aristotle described slaves) and dark prosthetic, became and merged with human being.
Sex, Race, and Robo-ethics
We should not forget that this historical transformation was greeted much like the Kurzweilian singularity, with the fear and fervor of apocalypse, particularly when it became clear that the political narratives and structures of racism were erected as much by scientific notions of difference as by the need to police desire. Perhaps of greatest significance here is the merging part and the elaborate sexual or reproductive mythos suggested by all talk of blending, assimilation, and socio-cultural intimacy. We have been here before. Levy recognizes this in his 2016 follow-up lecture, “Why Not Marry a Robot,” which quite naturally tilts his interest in human/machine intimacy and its subsequent cultural impacts and transformations towards miscegenation.4 This should be no surprise given the last generation or two of “cybertheory” and “cyborg feminism” and the emergence of “robosexuals” as a social type: one cannot fully contemplate human/machine hybrids without the process by which hybrids are generally made—sex.
“Why Not Marry a Robot” argues that the justification for human/robot marriage can be, should be, and ultimately will be sourced in interracial marriage. Central to Levy’s argument is the fact that the prohibition of such relationships was ultimately due to the belief that race implied a difference in biological status or species being, i.e., s/he was not human, or a legal person. Such prohibitions—as with same-sex marriages, he argues—will soon be culturally outgrown precisely in the way that anti-miscegenation laws and laws against interracial marriage have been. This use of race is not simply metaphoric, and not casual in the way that so many parallels between race, gender, and other forms of otherness can be. Interracial marriage is rooted primarily in the context of the earlier singularity, when blacks in particular were still considered inferior, at worst objects or animals and at best a lesser form of person. As an outcome of that singularity, interracial marriage signified a legal and moral redefinition of social and biological categories, which in Levy’s view will serve as a precedent in perpetuity.
Levy’s thinking is even more eccentric than this, and therefore more helpful in making sense of the increasingly difficulty to untangle historical and conceptual thickets of race, sex, and machines. His goal is not to just enshrine the human’s right to love or to see love as a primary sign of “the human,” but to protect and therefore assert the robot’s right to choose—or to be accurate, to if not choose, to not reject. “Humanness” is irrelevant here. Relation is what determines intimacy and what will demand social recognition. Intimacy will generate the claim on personhood. Intimacy is that most ethical and transformative of spheres and its range of affective realities are what shuffle and redefine social and biological categories via desire (i.e., by virtue of my love—and the power to legally render it so—this “thing” becomes human. It is worth keeping in mind that this merely restates something that we find more historically agreeable: by virtue of my disdain—and my power to legally render it so—this human becomes “thing”).
BINA48 (Breakthrough Intelligence via Neural Architecture 48), robotic face combined with chatbot functionalities, owned by Martine Rothblatt’s Terasem Movement, and modeled after Rothblatt’s wife. Hanson Robotics, 2010.
These debates may be less focused on robotics than Levy acknowledges, but they are clearly dependent on them in that they are rooted in historical and philosophical questions about “the human,” which are rooted in historical and philosophical questions about race. It is important to point out, though, that Western technology has always been haunted by the parallels that Levy so easily depends on, particularly between machines and slaves. It is a product of a drive to embodiment that is older even than modernity, but which came to cultural material fruition in the context of industrialism and chattel slavery. That very drive to manufacturing a future, one that depends on bodies and labor while promising escape from bodies and labor, ensnares technology in the problems of our past, and is why metaphors of race, sex, and reproduction entangle us in the material problems of imagining and inventing a future.
These days, however, the anxieties and arguments are primarily focused on algorithms, machine learning, neural networks, and their increasingly intimate social presence and cultural power. Though less focused on the anthropomorphism of previous technological anxieties, our moment has turned exactly to the sphere of ethics to engage or correct this legacy of human and machine. Problems, for example, now regularly appear in seemingly unpredictable spaces of pure invention—so-called misogynistic chat-bots in the virtual spaces of sociality, racist facial recognition systems, prejudiced algorithms in neural networks, the call for robot rights in advance of full automation, the specter of an AI that is human enough to be described as prejudiced.
And if this all seems still somewhat too abstract, there are the complex problems of gender, sex, and power at work in the emergent field and market of sex/companion robots, which will no doubt force a larger cultural reckoning with all these socio-political issues, particularly in the sphere of intimacy and, yes, sex and marriage. Suddenly, fields neither known or celebrated for their attention to race, sex, prejudice, or cultural sensitivity, such as computer science or robotics, find themselves ill-equipped to answer questions that have provided the subtext of much Science Fiction but which can best be identified when refracted through the lens of black and post-colonial thought, and feminism. Questions like, how do we in fact create synthetic bodies without the histories and habits of how we have previously treated, known, and organized different bodies? How do we apprehend a synthetic mind without tripping over the legacies of differential apprehension and the apprehension of differences? And what are the roles of race, gender, and desire in rendering object as subject, thing as being? We have been here before.
Personhood and Machine Autonomy
Machine learning, we know, often keys in on subtle excesses in our language and our patterns of representation, particularly those we are unaware of or silently acknowledge are inappropriate. In other words, machines make sense of our hesitations, know our secrets. We have long known that racism is often a change in breath, a type of hesitation, a sometimes barely recognizable response to certain bodies or unpredictable desires. This is how technology learns what we refuse to acknowledge about our past. What greater evidence of prejudice can there be than in finding it re-produced and amplified by a mirroring not of our naked prejudices but of our silences and denials, the imperceptible rustling of skin and breath? Until we find a way to program guilt into our technological systems as a way of monitoring or self-policing their responses, they will continue to be far more honest than we are about such information.
Lest we imagine these questions as purely abstract and in the private realms of individual ethics or paraphilia, corporate titans like Microsoft, Amazon, and Google, who now houses DeepMind Ethics and Society, and powerful groups like the Institute of Electrical and Electronics Engineers (IEEE), have begun to formally wed their interest in advancing technology to anticipatory or pre-emptive ethical standards. This is a response to those so-called “algorithms of oppression” or “weapons of math destruction” that suggest how surprisingly pliant our technologies are to human foibles.5 And just as machines and algorithms begin to display a likeness to humans via their tendency to discriminate, the European Parliament has even begun considering distinct regulations to govern them. Curiously, the regulation being considered is not focused on transforming design or programming parameters with an awareness of bias or a commitment to rooting out prejudice, or transforming the cultures of tech; instead the goal is to grant these technologies a distinct form of “electronic personhood” in lieu of full humanity. England’s Prime Minister, Theresa May, has for example declared that the UK should be the world leader in determining the ethics of Artificial Intelligence.
One wonders what this sort of “personhood” means since its focus is on liability and the reality of biases in algorithmic decision-making. Just who is responsible when an electronic person is judged to be racist or sexist or to have operated in discriminatory ways? And does not ethics imply some degree of reciprocity, or mutual recognition?
Sex dolls produced by the company RealDoll (California, USA), 2019
One cannot help but wonder if this personhood is less about ethics than it is a moral capitulation. Instead of addressing the socio-political and historical conditions of bias, it shifts the terms of debate in ways reminiscent of that old British imperialism which in the 19th century ended slavery primarily to use the ethical as a way to politically challenge the rising threat of America in the Atlantic region. The point was not to end racism but to use it to decry a political adversary. For example, Europe’s interest in morally regulating AI seems merely in anticipation of China’s stated determination to be the global leader in AI technology. We can see the old back and forth between types of imperialisms here—the one unabashedly technological, driven by profit motives and without any political restrictions, and the latter, in denial of its racial history and able to curb the former only via a new form of that soft, human rights imperialism that China has long accused the West of using to stunt Chinese industrial growth and globalization.
This global struggle over the ethics of machine intelligence should remind us that our futures with machines lie in personhood and its inevitable variants—not with clichés of an alleged “humanity” being usurped or romantic notions of its fundamentally untranslatable (or un-commodifiable) nature, or the kind of anxieties that dominated 20th-century culture and science fiction, those generated by anthropomorphism. The fact that “humanity” has been violently withheld from certain “genres” of humans, as Jamaican theorist Sylvia Wynter would put it, and that metaphysical, cultural, and economic systems have been erected on that denial, means that the term is far less absolute than its mythology insists.6 Personhood matters more because it has a much clearer history than “the human,” certainly in sexual and racial terms, and as a category has been far more socially flexible and legally mutable. To know who or what can or will become admitted into personhood simply requires a history of the category.
Can Objects Comply?
It is worth pointing out that the kind of personhood being discussed above is anathema to the high capital dependent companies that produce robotics and AI. Who would want to produce a commodity that you could not sell for fear of being charged with some form of trafficking, or a machine you could not employ without common restrictions of labor, which is to say, rights? Even manufacturers of so-called “sex robots” or “companion robots”—and Levy himself—are having to deal with anticipatory ethical standards, as well as organized protest framed in explicitly ethical terms and always vis-à-vis “real” women. But these are the primary cultural tensions and implications of “personhood” for manufacturers of machine subjects, or of machines that resemble or function as subjects.
After all, personhood was not only shaped within capitalist social relations but also specifically through chattel slavery. As a pliable and politically fraught category, it depends on animating the inanimate, or giving soul to the soulless. It features the bestowing of “life” or social recognition to objects, which then implies or threatens one with the possibility of intimacy—though it is more often the case that it is desire or intimacy that either breathes life into the inanimate or forces social recognition of the object. Here, one cannot help but reflect on critic/poet Fred Moten’s now famous declaration that the “history of blackness is testament to the fact that objects can and do resist.”7 Can we now query whether it is also true that objects comply, or can consent?
Take for example Levy’s discussion of the case of Angela Marie Vogel, the activist whose marriage to a corporation (or “corporate person”) was repealed in 2012 by the State of Washington.8 Even here, race becomes central to framing the possibilities of personhood as well as its legacy. The state argued that though endowed with the rights of a “natural person,” this “artificial person” could not marry because it was incapable of consent.
Oddly enough, the lack of consent in this case had little to do with the fact that the “person” in question was artificial. The real problem was the age of the “groom”: it/he/she was only a month and a half old, and the legal age of consent is eighteen. Statutory age notwithstanding, greater challenges will no doubt appear given the pliability of the category of “artificial person” and its validity as a legal strategy. Even though Vogel was making a legal point, challenges may arise from what is called “Objectum Sexuality” and related “objectophile” paraphilias, in which humans fall in love and/or form intimate, sexual, and meaningful relationships with objects and in some well-known cases, cars, bridges, and buildings (the Eiffel Tower, for example). They are most certainly emerging due to the growing market for “sex dolls” or “companion robots” and may yet put pressure on the legal system to recognize these relationships, provided of course that the “artificial persons” reach the legal age of consent.
Were Levy to dig a little deeper into American history and the history of chattel slavery, it would be clear that “Objectum Sexuality” and interracial marriage are not actually the primary precedents for human/robot love, nor are they the primary legal justification for the merging of flesh and fabrication. Both depend on something much more foundational and that historicizes interracial merging and paves the way for the inevitability of other such border crossings, that is, “corporate personhood.” As anti-corporate activist and writer William P. Meyers has famously written, “Slavery is the legal fiction that a person is property. Corporate personhood is the legal fiction that property is a person.”9 Both legal fictions emerge from the same place: the 14th Amendment.
Petrus Camper, Transition of the facial angle, from the monkey to Apollo, in Dissertation on the natural varieties that characterize the physiognomy (Paris: H. Jansen, 1791). Engraving.
Recall that the 13th Amendment abolished slavery in the United States. The 14th, adopted in 1868, was established to affirm equal protection under US laws for blacks as well as to grant citizenship rights to former slaves in the wake of the Civil War. It erased the notorious “Dred Scott” decision from 1857 which held that those descended from African slaves were ineligible for American citizenship, and paved the way for the 15th Amendment in 1870, which gave black males the right to vote. Due to an awareness of the still strong notion that blacks were not quite or fully human, as well as a great many social and cultural forces attempting to maintain some sort of racial hierarchy—including an enormous upsurge in racial violence—in the wake of the Reconstruction era, the Supreme Court made clear that the 13th–15th Amendments were specifically for “the freedom of the African race, the security and perpetuation of that freedom and their protection from the oppression of the white men who had formerly held them in slavery.”
Yet despite this codicil, the Court was unable to avoid the surprising elasticizing of personhood embarked upon by the Southern Pacific Railroad in 1881. The Railroad argued that special taxes on railroad property in California discriminated against it specifically under the 14th Amendment. Apparently, the law’s protection of racial identity could be manipulated to address corporate identity, since the very definition of “person” could now be challenged. In 1882 the railroad deceptively yet successfully argued that the drafters of the 14th had replaced “citizens” with “persons” in order to account for “artificial persons” as well as “natural persons.”
That personhood was or could be a state construct was clearly a product of chattel slavery, and the 14th Amendment attempted to rectify that primal dehumanization of Africans. In the context of law, however, it rendered “personhood” far more fluid than nature or history had previously allowed. This redefinition of personhood would benefit corporations far more than those it initially intended, because by 1896 the Court legalized segregation via Jim Crow laws with the notorious Plessy vs. Ferguson/separate but equal doctrine whilst corporate culture would only expand its influence via that same legal logic of “artificial personhood.”
New Species, Old Prejudices
And so, we end at the beginning. In historicizing the relationship between ethics and technology, humans and machines, we wind up at that precise historical moment where the first singularity was ratified. Here, personhood was first deployed to rectify a historical ill (dehumanization) by way of legally constructing a new kind of human being. It then revealed the category as fluid, thus opening the gates to expanding personhood in inorganic, “artificial” terms, something clearly crucial in the movement towards a recognizable if not fully autonomous AI. But given the long historical imbrication of race and technology and the inescapability of one history and sphere of meaning from the other, we also end at the future, the precise location where that next singularity is already being naturalized and may one day be justified.
Paul Richer, Male Model, Venance, in «Carnet de fiches anthropométriques», 1923-1939. Paris, ENSBA
So, what now?
Well, for one thing, the impact of racial, sexual, and cultural differences must be taken seriously as we acknowledge the near-inevitability of artificial beings and some forms of Artificial Intelligence. From this position between singularities, it should be clear that our social acceptance and cultural tolerance for such beings will likely have little to do with how “human” they seem or how like us they may be. What will matter far more is how we make sense of them via histories of otherness, from subservience and domination to anxiety and pleasure. As suggested by Despina Kakoudaki, we might finally be ready to push beyond the aesthetic and cultural limits marked by the Freudian uncanny because “verisimilitude” is beside the point.10 For example, we have learned that robots that look too much like human beings can trigger a hostile and suspicious response, but those that are deliberately modeled to look like characters or exaggerations, or vividly inhuman, can actually generate great empathy, warmth, and affection. We also know that feigned affections can generate authentic attachments. Therefore, so can programmed ones, especially if we are exposed to them as children, which is the point of the “sociable” or “domestic” robots produced for use with children or the elderly.
Again, what will matter more is not what machines look like but how we interact with them despite visual or epidermal differences, or despite even glitches that will inevitably be read as either signs of personality or the equivalent of cultural variations. Of course, when it comes to algorithms or an utterly de-anthropomorphized or non-representational computer landscape, none of this will matter at all. But socialization, power, and intimacy will still occur despite the fact that the true agency that we ascribe to cognition and “free will” may either never actually be constructed or capably recognized. “Free will” or autonomy will have nothing to do with our social, political, or sexual recognition of machines considering how historically contingent such things have been with regard to, say, blacks—who without them still functioned in the most intimate spaces of white life during slavery. And in the wake of legal personhood, blacks have had their “humanity” relentlessly questioned because in truth, that personhood was meant to maintain a distinction from a full humanity considered to be the sole property of whites.
Ultimately, the social recognition of robots or AI will have everything to do with how they function in relation to humans, as well as how they are governed by a set of common and reciprocal values generated to contain the raw fear of otherness while policing the terms by which intimacy is possible—providing, of course, that machines are capable of honoring those values and their capacity for reciprocity is acknowledged. Either way, race will continue to be decisive in the cultural and legal process of making machines persons, if not human, because the quest for an ethics of machine/human relationships and the drive to inculcate or program an ethical matrix into technology is dependent on the history here discussed. Not only does this history guide our metaphors, and haunt rules of representation, it should continue to. This is because it conditions how we can even conceive of relationships founded on differences so radical that they were historically imagined and legislated as differential between species, or between humans and objects.
Because it is the historical sublimation of slavery, immigration, and their attendant anxieties that ultimately accounts for the contemporary attempt to create an ethics in public relationships with AI, it should also be the case that an honest reckoning with the last singularity and its legacy will generate the terms by which such beings will be engaged. This is the case from miscegenation to cybernetics, and from immigration to assimilation and Civil Rights. Science Fiction, in fact, tells us that it is the very refusal to change our historical relations with “the other” that often inspires those hackneyed but enduring “revolt of the robots”/rise of the machines narratives in the first place. So, given the nature and long impact of that last singularity and the fact that its echoic traumas forever color conceptions of life itself, machine intelligence and artificial life will explicitly depend on extant histories of cultural otherness—at least in terms of personhood, but at most in terms of citizenship and its relentlessly reciprocal demands.