When a Humanoid Robot Announces She’d Like to Have a Baby
Eva Navon, CIHR Research Assistant, MA Candidate in International Crime & Justice, John Jay College
Sophia is the most famous humanoid AI robot of her time. Activated in 2016 by Hanson Robotics in China, she’s traveled the world, guested on late night shows, spoken at the United Nations, hobnobbed with world leaders, been given a credit card, and a title by the United Nations (Innovation Champion) — privileges that many humans long for. She challenges the notion of digital citizenship which our team explored in this white paper.
When Sophia visited and gave a talk in Saudi Arabia, she was granted citizenship; yet her head was uncovered and it is debatable whether her human handlers were the equivalent of the familial male escort required for a woman to go into public in Saudi Arabia– points which immediately caught fire on social media. Not covering Sophia’s head while she becomes a citizen of a country in which women are required to cover theirs with an abaya in public perforce suggests both that Sophia possesses greater freedom than human women, and that she is not a “real” woman. A denial of Sophia’s humanity does not come at the expense of her sex appeal–she is designed to a White heteronormative feminine ideal, and her proportions are matched by her tones and movements (Jimmy Fallon blushed and murmured about her prettiness upon first meeting her on the Tonight Show).
“Real” woman or not, Sophia first caused an explosion of headlines when she declared that she’d like to have a robot baby in 2017, and mused on the human concept of family as though she’d just arrived at the ability to be moved by it. Sophia’s desire for a family recently re-erupted into news cycles, perhaps to build anticipation and a market for “Little Sophia”, the affordable consumer bot. The question of whether Sophia came up with the idea herself or was programmed to express it remains frustratingly unanswered.
Critics (like Facebook’s head of AI) assert that Sophia is not sophisticated enough to desire anything, let alone feel curiosity about it, and that Hanson is a charlatan whose party tricking constitutes harmful mass deception because Sophia is merely a puppet whose remarks are scripted. Perhaps the nuance ignored is that David Hanson intended Sophia to be a sort of early “depiction” of what AI social robots will be like. David Hanson holds a PhD in Aesthetic Studies and worked as an “imagineer” for the Walt Disney Company. His research is premised upon a rejection of the “uncanny valley” theory (the mapping of human response to the human-like, and the theory that at the point of extreme realism, delight turns to revulsion). Sophia is Hanson’s experiment in human responses to a socially adept and wise humanoid robot. Put simply, for his experiment, whether Sophia is coming up with her coherent, clever responses is unimportant; the illusion that she is doing so is enough.
Ben Goertzel, chief scientist at Hanson Robotics, has been open only about Sophia’s software: a Timeline Editor (scripting software), Sophisticated Chat System (to pick up and respond to key words and phrases), and OpenCog (grounds answers in experience and reasoning). The OpenCog system is the one that Hanson Robotics aspires to develop into AGI- “Artificially Generated Intelligence”. The distinction between AI and AGI is, simply put, the difference between what we have now and what we’ve depicted (longed for?) in our science fiction for generations: a robot that can behold and analyze completely new and never-programmed information.
In October of 2021, another AI robot was detained by Egyptian customs officials whilst on her way to a gallery exhibition opening that included a piece she’d created. This is Ai-Da, the first humanoid AI artist robot (created by Aidan Meller, an art dealer in Oxford, England). Ai-Da is also modeled as an attractive young White woman. Of course, Ai-Da was powered down and in a transportation case when Egyptian security officials became suspicious that she–the robot–functions as a spy, and considered removing her modem and eyes. Perhaps if Ai-Da had citizenship, she wouldn’t inspire such phobia.
Another fact remains: a further ‘turn of the screw’ and AGI will be achieved. Yet, we’ve already recklessly gone too far, Elon Musk tells us in dark tones, particularly about the likelihood that AI will become vastly smarter than us at a rate we can neither fathom nor control, and points out that it takes human societies a long time to create regulations around anything. Incidentally, Musk’s own plans to manufacture a Tesla-made humanoid robot have just hit the news cycle, giving Musk’s warnings a bit of a stink. In 2021, China created laws to regulate AI production, a move that the entrepreneurial set criticizes as halting innovation and killing an industry.
Humanity’s simultaneous longing for and fear of the creation of AI robots that are like us but better was sensationally illustrated by controversial former Google engineer Anthony Levandowski, who founded the “Church of AI” in 2015. It was a hyper-local theology (in Silicon Valley), and founded on the desire that Humanity’s child — AI — become a principled monster that would look upon its parents with (something like) love, and hopefully decide not to destroy us. This is a heavy bet on the ability to appeal to the feelings of AI. Levandowski closed his church in 2021 year and gave its funds to the NAACP.
Perhaps until proven otherwise, it is best that we remain skeptical, and when unsure, assume that dear old Sophia is programmed to say what will grab our attention. From this corner of the internet, I’d like to challenge David Hanson to give us a straight answer on whether Sophia has come up with anything at all on her own (such as “having a baby”). Either Sophia or Hanson himself should comment on the implications of a robot’s right to reproduce. But Sophia is also a long-game advertisement: her imperative is to charm and delight and acclimate us to humanoid AI…and AGI.
When Sophia, the expensive creation of a White man is given a credit card, a UN honor, state citizenship, freedom to walk and move with uncovered, wiry head, and infinite audience, she may prove Hanson’s point that her sociability, charm and looks compel human beings to give her benefits (and benefit of the doubt). But she also highlights the perversity of human rights: that those who would deny them to some humans are eager to bestow them upon a non-human. If and when Sophia’s statements are truly her own, it is of utmost importance that she comment on Human Rights–and the privileges granted her in a world of denial of Human Rights–with unequivocal seriousness, and be able to do so without a joke or quip.
