Can a Robot Join the Faith?

The recent awarding of Saudi citizenship to an android named Sophia raises a strange and unexpectedly pressing theological question.

As an apparent coup d’etat ripples through Saudi Arabia, the rising ruling faction is trying to keep things upbeat by sending bullish signals to the world’s mega-rich. Exhibit A is Neom, part of the kingdom’s Vision 2030 initiative, a proposed utopian city whose modest slogan is “the world’s most ambitious project.” Neom imagines itself a swinging, sort-of-liberal international trade center, built from scratch, at a cost of five hundred billion dollars, on the shores of the Red Sea. According to its official Web site, Neom will be an “aspirational society that heralds the future of human civilization,” which means, of course, that it will be operated and inhabited by armies of artificially intelligent bots. As part of the rollout for Neom, the Saudis have just granted official state citizenship—a first for planet Earth—to one such machine, named Sophia Robot.

Sophia is a chatty A.I. android born of the Hanson Robotics lab, in Hong Kong. In conversation, she will look you in eye in order to memorize your face; she will present, in succession, like cards from a deck, her sixty-two facial expressions; she will wink in patterns designed to put you at ease. Sophia processes speech and, above all, she learns constantly. Her creator, David Hanson, an alum of Disney, is the kind of person who throws around such phrases as “framework for computational compassion.” The surprise announcement of Sophia’s new Saudi citizenship was made at the Future Investment Initiative conference, in late October, where she was interviewed onstage by Andrew Ross Sorkin, of the Times. In her remarks, she flattered her audience as “smart people who also happens [sic] to be rich and powerful.” With comically horrible comedic timing, she cracked jokes about Elon Musk’s anti-A.I. warnings; she even managed to output some witticisms about not destroying humankind. (That interview, and others like it, are believed to be scripted, and they certainly come off that way.)

Amid the oohing and aahing about Sophia’s uncannily lifelike appearance—she was modelled after Audrey Hepburn, an homage slightly complicated by her built-in neck zipper—there were several objections. Making Sophia a citizen, some commentators noted, effectively gave her more rights than most Saudi women. It was also an insult to the kingdom’s minority groups, especially to migrant laborers, who have been denied citizenship for generations. An article in Newsweek suggested that Sophia, as a non-Muslim, was not even technically eligible. This claim, it turns out, is not quite accurate: there is no explicit mention of religion in the country’s immigration regulations, though officials may factor it in at their discretion. (A more pertinent legal question is whether Sophia can prove Saudi residency for the past five years.) But what about the premise—that Sophia Robot might, in theory, be considered Muslim, or Christian, or Jewish, or Buddhist, or Jain? Could Sophia, or somebot like her, be raised within, or convert into, and practice a faith? And, if so, could a robot serve as a rabbi or imam, or as Pope?

Some theologians might argue that even an advanced android might, by definition, have trouble crossing over into the realm of Homo religiosus. For a Christian thinker such as Kierkegaard, the facts of religious truth are supported not by proof but by passion. Reason and logic, and the ability to improve in those capacities, can be hindrances to true belief. A leap forward in intelligence, especially of the machine-learned variety, may render the Kierkegaardian leap of faith that much more difficult, if not impossible.

For others, especially for people of law-based religions, the stakes are more practical and quotidian. When I was a student, the edgy rabbi at my yeshiva introduced me to a legal opinion penned by Tzvi Ashkenazi, the seventeenth-century rabbi of Amsterdam, on the question of whether a golem—an animate clay man—can be counted in a minyan, the quorum of ten required for the thrice-daily communal prayers. Ashkenazi cited the case of his own illustrious great-great-great grandfather, who had created just such a being. (Later, Ashkenazi’s son elaborated that, when this ancestor tried to deactivate the golem, it attacked him.) Ashkenazi ends up mostly unconvinced of the golem’s claims to membership in the Jewish community—but, in a touchingly humane turn of thought, he also acknowledges that a golem is a kind of orphan whose status may well depend on who raises it. The door to religion is thus left open to the orphaned androids of the world, creating the very real possibility that, in the near future, the girl celebrating her bat mitzvah by doing the robot on the dance floor may actually just be an awkward adolescent robot who is trying, and failing, to dance like a human.

Some Muslim writers share this emphasis on the legal and social aspects of the issue. About three hundred years before Ashkenazi pondered the golem, the Muslim scholar Muhammad ibn Abd Allah Shibli took up the question of whether, according to Sharia law, a jinn, or spirit, could marry a human. Shibli cited historical examples of humanoid jinns marrying humans; for him, the question wasn’t whether such a union was physically possible but whether it was permissible or forbidden, halal or haram. Like Ashkenazi, Shibli tilted his judgment toward haram, but he, too, acknowledged that the debate remained open.

But, if you think about it, the question predates the medievals. The Biblical account of creation, in the second chapter of Genesis, can be read as an A.I. project gone awry. According to that account, “the Lord God formed a man from the dust of the ground and breathed into his nostrils the breath of life, and the man became a living being.” Notice the order: first the thing is given shape (dust for Adam; Frubber for Sophia), then it is animated with life. Only after that happens is it endowed, in the verses that follow, with intelligence. The creator programs it via commands, inputting what to eat, what not to touch, what kinds of knowledge to avoid. As we all know, those first two A.I. units, designed to be learning machines, soon began to think outside of their intended parameters. Can bots go haywire and commit mass genocide, launch nukes, enslave other bots, and wreck the biosphere, without pity? This nightmare has been happening forever. It’s called human history.

Whether robots can join a particular religion may be secondary to the question of which religion they are already being programmed into. Perhaps you’ve heard of the apocalyptic cult, cultivated among technologists, that centers on the doctrine of the Singularity—the belief, as Jaron Lanier once described it, that “one day in the not-so-distant future, the Internet will suddenly coalesce into a super-intelligent A.I., infinitely smarter than any of us individually and all of us combined; it will become alive in the blink of an eye, and take over the world before humans even realize what’s happening.” (At a recent Singularity summit, it was prophesied that this Rapture-like moment would happen around 2040.) Whether you see the A.I. end-times as a triumph of evolution or the start of a bitter era for humankind all depends on your outlook. But when the Singularity comes to pass, and the bots write their own bibles, the more pressing question may not be whether bots can join our religions but what place we will have in theirs.