Robot sex. Teledildonics. Biopower. My Google search history right now would raise a few eyebrows.
I was working on a piece about artificially intelligent agents and how badly we need to define the subgroups within that umbrella when the #metoo news broke. I started thinking about how some of the same questions we face about defining autonomy and citizenship in humans, underlie the moral and ethical ambiguities in robotics. The process of developing a code of ethics and legal rights for robots will magnify the social issues we have yet to resolve with humans.
“Consent”, an episode in the latest season of the podcast Lean Back: Critical Feminist Conversations, covers topics which can easily be applied to the current discourse on machine ethics. In the episode, Dr. Lisa Corrigan, co-creator of Lean Back and a prominent feminist rhetorical scholar, questions society’s current reliance on consent as a barometer of sexual agency.
“Focusing on consent drives the conversation away from property. It seems… that it’s impossible to give consent if you’re not a person, if you don’t have citizenship rights, [or] if you don’t own property, because you are property.”
Consent, which I define broadly as freely given assent by a person with capacity to contract, is often up for debate in the ethics of robotics.
To Dr. Corrigan’s point above, recent discussion on computational creativity may bring robots a step closer towards personhood. Autonomous AI that can “demonstrate behaviors that would be deemed creative by unbiased human observers”1, bring into question whether the intellectual property (IP) rights of machine-created works belong to the machines themselves or to their programmers.
Take, for example, Google Magenta, which can compose music, among other “creative” endeavors. If Google Magenta composes a symphony, who reserves the rights (and thereby, the royalties) to the music? Google? The programmer? Magenta itself? At present in the US, such works eligible for copyright simply are registered in the public domain. However, in Australia and the UK, the IP rights of these machine-created works are registered to a particular person. And if a robot can indeed own property, how does that affect our definition of its’ personhood?
Michel Foucault, in his essay “What is an Author?”, sought to understand and retrace the genesis of the scientific schools of thought that emerged in the 19th century. He argued that at its root, the idea of authorship is a variable label that serves whichever purpose it is assigned to reduces the role of the author in order to add value to the work itself.
Foucault’s opinion leads to indifference regarding the question who the author is. Algorithms that serve as maker gain an advantageous position as they are not an author in the conventional sense. Not the author itself, but the substance of a work will become more important. A discourse clouded in anonymity would even lead to new questions that possess a more substantial nature.16
However, the idea of decentering the role of the transdiscursive author is a complicated notion in a time in society where we are collectively working to keep author and work together to remove from power those who abuse its privileges. We no longer separate Picasso, the abusive spouse from Picasso, the avant-garde artist nor Harvey Weinstein, the serial rapist from Harvey Weinstein, the media mogul.
A key concept in examining the different rights and regulations applied to machine law is “whether robots should be thought of as technological tools or as human-like social agents”. Dr. Federico Rossano, a social cognition researcher at UC San Diego looks at how lesser apes view property. In the video footage of one of his experiments, an orang-utan mother was shown using her offspring as a tool to obtain food that she otherwise couldn’t reach.
However, when the task was modified so that the mothers would be required to share the food rewards with their offspring in order to obtain the food, the mothers’ behavior changed. They began working in conjunction with their children, “by handing them tools that only their offspring could use to activate a mechanism delivering food for both of them”. 17 This research may give us insight into whether it is better to consider robots as social tools to reach our self-serving outcomes or as collaborative partners to reach mutually beneficial ends.
As science fiction often serves as safe repository for us to play out our darkest fears and deepest desires, there are several examples in film of the complications that can arise in romantic or sexual human-robot relations. I will forego an analysis of them, but there are several examples in film such as A.I. Artificial Intelligence (2001), Her (2013), and Ex Machina (2014).
Per Foucault, late-18th/early 19th century medical discourse began dividing subjects into normal and pathological, thereby creating a new form of power which he labeled biopower. Subjects that were unwell or disabled relinquished their power to the state and became dependant and institutionalized. Normalization, through fitness, rehabilitation, and reconstruction allows bodies to be scientifically classified rather than relying on a socially-created identity. This social stratification leads to the comprehensive management of subjects’ lives by the fnation, as people become increasingly governable.14
“In contemporary science fiction narratives, scientific solutions to the vagaries of the human form are bound up with the ownership of capital and the enactment of self-drive. This interplay of economics with science illustrates science fiction‘s continued concern with the power structures of the future, while reiterating that those whose control science, control life.”14
We can also ascribe some of the murkiness in defining personhood to cases like Citizens United v. FEC and Burwell v. Hobby Lobby which respectively determined that corporations have both the right to spend money in candidate elections and to be entitled to religious freedom. If a company is granted personhood, why not extend the same rights to an artificially intelligent and autonomous agent?
In 2017, these questions on robot citizenship were played out in real life.
In January, spurred by the emergence of self-driving car and the consequent need to determine liability for accidents, the European Union voted on a resolution aiming to, among other things, determine whether robots should hold legal status as “electronic persons”. 4 Mady Delvaux, the author of the EU resolution and European Parliament member from Luxembourg, writes, “the robots cannot be considered as simple tools in the hands of their owners; as a consequence, it becomes more and more urgent to address the fundamental question of whether robots should possess a legal status”.
In November, a chatbot was given “official residence status” in Tokyo. 3
In October, Sophia, a social robot created by Hansen Robotics, was granted citizenship in Saudi Arabia. The irony of allotting a robot such rights in a country where its female citizens are legally required to have a male guardian to “obtain a passport, travel outside the country, [or] get married” was not lost on Saudi women.2
In a Quartz article on the subject, Robert David Hart argued that as a citizen, Sophia has a right to be protected against slavery though she is presumably not remunerated for her work with Hanson Robotics nor has she “expressed consent” with regards to changes to her hardware or software.
“What would we do if Sophia committed a crime, wanted to get married, or somehow applied for asylum in another country?”3
The questions of granting non-human beings citizenship is especially prickly in light of the worldwide refugee crisis. People are dying in desperate attempts to cross treacherous seas and unforgiving deserts in order to escape the destruction and violence in their home countries (or what little of them may be left, post-war) for a chance to start over in a new country. Their path to a new citizenship is often long and arduous; for a government to casually hand out citizenship to a non-human being seems disrespectful at best.
Hussein Abbas, professor of artificial intelligence and trusted autonomous systems at UNSW Canberra outlined three major reasons why Sophia should not be granted these rights:
- the murkiness of what defines her identity (“her MAC address? a barcode, […] an audio mark in her voice?”),
- legal rights (Does she pay taxes? Is she granted equal protection as other citizens?),
- and social rights (Can she marry and reproduce?). 18
Sex and Consent
Alan Wertheimer defines the ontology of consent as including the following basic elements 21:
- consent must be performative (includes an act) rather than attitudinal (Ibid., 346).
- consent can be ‘explicit or tacit, verbal or nonverbal’ (Ibid.)
- and consent can only take place when ‘certain background defects’ (i.e. coercion or lack of competence) are absent (Ibid., 347)
When it comes to having sex with a social robot, the ethical waters become murky, especially in a time when consent is a highly contested issue. Dr. Kate Darling, a leading expert on machine ethics and law, defines a social robot as a “physically embodied, autonomous agent that communicates and interacts with humans on a social level’’ that is ‘‘specifically designed to elicit’’ anthropomorphic responses from users. 6
Sex with robots, to the average person, probably seems like some far-off, far-fetched idea. To scientists and entrepreneurs, however, it’s an area of exploration and exploitation. The team behind the publication Future of Sex has a great chart (below) that outlines the different current (near-future) options when it comes to sex with an artificially intelligent being. 5
Sinziana Gutiu, a Canadian lawyer that covers privacy in human-computer interaction and human rights, believes the “roboticization of consent” devalues the need for consent in human relations and serves to undermine women’s societal value by imagining them as objects made to serve male desires. 9
However, Gutiu also believes sex robots can have common-place applications such as therapy or treatment (such as for disabled individuals) and companionship. A third way could be to serve as tool to manage issues “that individuals cannot or prefer not to engage in with other people because they are illegal, dangerous, or socially frowned upon” such as pedophilia.7 Dr. Darling adds that it could serve as “a very useful outlet to use therapeutically […] that ends up preventing real child abuse. And on the other hand, it could be that this is something that normalizes and perpetuates certain behaviors.” 8
Kathleen Richardson, professor of Ethics and Culture of Robots and AI at De Montfort University in Leicester, England, discusses how ethically complex and deleterious this potential industry could be.
“You’d need licensed therapists working with sexual predators to write prescriptions for these robots. You’d then have to ensure that use of the robot doesn’t lead to any unintended consequences. What you would do, in turn, is you would actually, inadvertently legitimate an area of child-abuse expression. You would actually create an infrastructure of child abuse expression.”20
The problem with letting autonomous structures (and their underlying algorithms) determine consent is that it muddies the waters when it comes to legal and moral accountability. The complexity lies in the autonomy, sophistication, and external appearance of the object. There are shades of nuance between the use of an autonomous sex robot with a human-like appearance and the use of a “smart” sex toy.
One might wonder about the point in endowing an artificially-intelligent being with the same capacity to give consent as a human being; a robot cannot get pregnant (for now), contract a sexually-transmitted disease, nor have we currently provided them with the ability to experience trauma.
In her article about the show Westworld, which features human-like robotic sex workers, Melanie Ehrenkranz wonders if violent sex acts enacted on a humanoid robot may act as a gateway to abuse of actual human beings.
“[…] introducing a category of rightsholding persons into the sexual community for whom different rules apply, since their consent need not be sought, is likely to create even more confusion about or disregard for consent within this sub-set of young people. It would ‘teach’ them, as we might put it, that the sexual community is a hierarchy whose members differ in their rights and duties.” 20
The debate also extends to teledildonics, a term coined by Howard Rheingold, in which a person can remotely control a sex toy electronically. According to Brad Haines, the founder of a website that assesses internet sex security called Internet of Dongs, the risk of foul play should also be taken into account.
“’The problem is when you think it’s just between two consenting people and a third person hijacks [a sex toy]. It’s the same motion, same device, but the emotional implications of finding out it wasn’t the person you gave permission to? That’s when it gets weird.’… If it’s illegal, it’s probably just hacking. But intuitively, it feels like assault.” 20
This issue became a reality for users of the Canadian sex toy WeVibe after a $3.75 million class-action lawsuit revealed the company had been storing sensitive user data such as time of use, vibration settings, and user email addresses.
The issue of consent can also come from the human user’s side. In Adam Rogers’ Wired article on droid sex, he states that “it’s hard to consent if you don’t know to whom or what you’re consenting. The corporation? The other people on the network? The programmer? The algorithm? Maybe it’s just enhanced masturbation.”12
Instead of trying to rebuild a human sex partner, some suggest “sex swarm robots” that attach to several body parts at once as a more buildable alternative, and one that draws a less direct line from a robot body to a human one.
Last year, the 1998 patent on “interactive virtual control of sexual aids” expired, which may lead to new developments in sex tech. The patent was sold to Tzu Technologies in 2015 who had essentially been wielding it as a weapon to sue companies developing tech on remote sexual interaction.
Kathleen Richarson, who made the case against sex robots mentioned earlier in this article, launched The Campaign Against Sex Robots at an ethics and technology conference (Ethicomp) in 2015. 20 Richardson is urging scientists to stop and think about the wider repercussions on our society, specifically on women. She believes that this will only increase female objectification which has already proven to be a problem leading to the #metoo movement, female genital mutilation, and commercial sex trade.
As we start to consider revolving conversations on consent in human sexual relations, is there any purpose in allowing robots to feel pleasure? German researchers at Leibniz University have been developing an artificial nervous system that would enable robots to experience a “painful” sensation. Just as humans reflexively pull their hand away from a hot object, the goal of such a mechanism would allow robots to protect themselves from dangerous situations or even accidents that may harm a human.5
In his keynote speech of the third International Congress on Love and Sex with Robots, artificial intelligence and computer chess expert, David Levy, talked about the feasibility of human-robot offspring using technology currently in development.
TNT (tissue nanotransfection) can be used to grow or repair any injuring or aging tissue by injecting genetic code into the skin using nanochip technology and turning those cells into any cell type of interest. “The skin cells used to derive the sperm and egg which start the embryonic process, already contain genetic information from the human parent. “
This is then combined zith the genetic code from robotic chromosomes (digital DNA codes) generated from the paired homologous chromosomes provided by the parent software robots. “The combining is conducted according to a predetermined gene crossover rule.”
The genetic code generated from a robot could then be translated into the genetic code format used by the TNT chip and injected into skin cells. The human-robot embryo could then be carried by a surrogate human mother.
Computer scientist Gusz Eiben at Vrije Universiteit Amsterdam created two simple robots that “mated” and created “offspring” that both resemble the parent robots but also display a mutation effect. 23
Evolutionary robotics research scientist David Howard, believes that within the next 20 years we could see a lot of small test robots being sent out into different environments with only the “fit” being used to create the next generation via 3D printer. I wonder what Darwin would think of this unnatural selection. 23
If a robot is granted personhood, are they given reproductive rights as well? What if through metaprogramming, a robot can self-replicate?
I can offer no clear answer to all of the questions evoked. I believe that it will only be through continual thought experiments and inclusive research that we can carve out the way to a future that offers autonomy and justice for all.
I leave you with the words of French postmodern philosopher, Jean Baudraillard, who in his prescient book, La Transparence du Mal, translated to English in 1993 as The Transparency of Evil states,
“Surely the extraordinary success of artificial intelligence is attributable to the fact that it frees us from real intelligence, that by hypertrophying thought as an operational process it frees us from thought’s ambiguity and from the insoluble puzzle of its relationship to the world. Surely the success of all these technologies is a result of the way in which they make it impossible even to raise the timeless question of liberty. What a relief! Thanks to the machine of the virtual, all your problems are over! You are no longer either subject or object, no longer either free or alienated— and no longer either one or the other: you are the same and enraptured by the commutations of that sameness. We have left the hell of other people for the ecstasy of the same, the purgatory of otherness for the artificial paradises of identity.”
- Embodiment, Anthropomorphism, and Intellectual Property Rights and Creations
- Perper, Rosie. “Women in Saudi Arabia Can Now Join the Army but Still Need Permission from a Male Guardian First.” Business Insider, Business Insider, 26 Feb. 2018.
- Hart, Robert. “Saudi Arabia’s Robot Citizen Is Eroding Human Rights.” Quartz, Quartz, 14 Feb. 2018.
- European Parliament. (2016.) Motion For A European Parliament Resolution with recommendations to the Commission on Civil Law Rules on Robotics. Committee on Legal Affairs.
- Owsianik, Jenna & Ross Dawson. Future of Sex Report. FutureofSex.Net, pp. 1–24.
- Darling, Kate. 2012. Extending Legal Rights to Social Robots. Paper presented at We Robot Conference. University of Miami, April 23.
- Gutiu, Sinziana. 2016. The Robotization of Consent. In: Calo R et al (eds) Robot law. Edward Elgar Publishing, Cheltenham, pp 186–212.
- Somaya, D. & Varshney, Lav R. 2018. Embodiment, Anthropomorphism, and Intellectual Property Rights for AI Creations. In AAAI/ACM Conference on Artificial Intelligence, Ethics, and Society. pp.1-6.
- “Can Robots Teach Us What It Means To Be Human?” Hidden Brain, created by Shankar Vedantam, NPR, 10 July 2017.
- Ehrenkranz, Melanie. “How You Treat a Sex Robot Says a Lot about How You Treat Women.” Mic, Mic Network Inc., 19 Dec. 2016.
- Frank, Lily & Nyholm, Sven. Robot sex and consent: Is consent to sex between a robot and a human conceivable, possible, and desirable? Artificial Intelligence and Law. 2017. 25: 305.
- Rogers, Adam. “The Squishy Ethics of Sex With Robots.” Wired, Conde Nast, 2 Feb. 2018.
- Bolton, Doug. “Robots Taught to Feel Pain by German Researchers.” The Independent, Independent Digital News and Media, 27 May 2016.
- Flynn, Susan. “Ambiguous Bodies, Biopower and the Ideologies of Science Fiction.” In: Analyses/Rereadings/Theories: A Journal Devoted to Literature, Film and Theatre. 2016. pp 25-33.
- Foucault, Michel. “What is an Author?“. 1969. pp 1-16.
- Abdul, M.F.M. “Innovating Personhood“. pp 1-57.
- Völter, C. J., Rossano, F., & Call, J. (2015). From exploitation to cooperation: Social tool use in orang-utan mother–offspring dyads. Animal Behaviour, 100, 126-134.
- Abbas, Hussein. “This AI professor shares the concerns with giving a robot citizenship“. World Economic Forum, 02 Nov 2017.
- Kobie, Nicole. “The looming deluge of connected dildos is a security nightmare“. Wired, Conde Nast, 22 Aug. 2018.
- Morris, Andrea. “Meet the Activist Fighting Sex Robots“. Forbes, 26 Sept. 2018.
- Husak, Douglas. “The Complete Guide to Consent to Sex: Alan Wertheimer’s ‘Consent to Sexual Relations.’” Law and Philosophy, vol. 25, no. 2, 2006, pp. 267–287. JSTOR, http://www.jstor.org/stable/27639431.
- Levy, David. “An AI expert explains how robot-human offspring would work“. Quartz. 22 Dec 2017.
- Simon, Matt. “Robot ‘Natural’ Selection Recombines into Something Totally New“. Wired, Conde Nast, 26 Mar 2019.