THREE

FEMALE ROBOT CAREGIVERS, DOUBLES, AND COMPANIONS

IN THE BRITISH AMERICAN TELEVISION series Humans (2015–18), a retired artificial intelligence researcher, Dr. George Millican (played by William Hurt), is plagued with memory loss and the side effects of a stroke. He is helped by a “synth” (short for synthetic) caregiver named Odi, a robot manservant. But Odi is malfunctioning, and authorities send an updated synth replacement, a stern Gestapo-looking woman, Vera (played by Rebecca Front), who makes Millican’s life a misery. She orders him to change his clothes, take his meds, and do his exercises. Like a drill sergeant, she is unrelenting in making sure he does the right thing.

In a familiar female role as caregiver, she nevertheless shows no sign of sensitivity or warm empathy and makes Millican long for his male synth Odi, who had far more sensitivity and gentleness but was malfunctioning. Millican says to Vera, “You’re not a carer. You’re a jailer. Get lost!” adding, “I don’t take orders from you!” Throughout, Vera as robot is always mechanistic and calm, a comic version of a familiar female paradigm—the nagging scold who won’t let Millican be his own man. Eventually, he locks her in a closet as he tries to make his escape.

As a series, Humans presented a thoughtful look at the implications of a world increasingly infused with simulations created to offer comfort, provide sexual pleasure, and help humans—with all the acceptance and resistance that entails. Adding layers of complexity to conventional paradigms of the female caregiver, wife, mother (see chap. 4), the series sometimes satirized, sometimes championed, the myriad roles of women in an increasingly technological world. It was also very much in touch with contemporary developments in robotics, particularly the quest to create empathetic robot caregivers.

Indeed, contemporary female robots not only are serving as sexy pleasure dolls but also may be designed to—eventually—provide health care and companionship for the elderly and the disabled. In America and countries including Japan, the age group of sixty-five and older has been the fastest growing segment in the population, and there will be an increasing need for professional caregivers. Foreseeing that there may not be enough qualified human helpers, roboticists at research centers, including MIT, and international manufacturers are scrambling to fulfill this need, creating “social assistive robots.” The robots are also being designed to be helpful for people with Alzheimer’s disease and dementia.¹

Professor Nancy S. Jecker of the University of Washington has been a strong proponent of caregivers and companion robots, arguing that they can play an important role for socially isolated people and people with disabilities. Social isolation and loneliness experienced by older adults can have a negative impact on their health, and, as Jecker wrote, they are factors “associated with a greater incidence of major psychological, cognitive, and physical morbidities.”²

Researchers have reported that patients responded well to robots that have humanoid voices and appearances, and some roboticists have gendered this concept, suggesting that they are creating their pretty, ultrarealistic female robots not only to serve as receptionists and even actors but also to serve as healthcare providers. David Hanson of Hanson Robotics in Hong Kong and Professor Hiroshi Ishiguro, director of the Intelligent Robotics Lab at Osaka University in Japan, developed attractive robots that usually look like young women in their twenties with expressive faces and rudimentary interactive speaking capabilities.

Since 1995, Professor Ishiguro has devoted much of his career developing both male and female humanoid robots such as the female robot Repliee Q2, the female androids Otonaroid and Kodomoroid, and he and his colleagues worked on creating the aura of empathy in robots like Erica introduced in 2017. Professor Nadia Thalmann, director of MIRALab at the University of Geneva in Switzerland, has for several years been developing her robot Nadine as a companion for children and the elderly. The robot is modeled to resemble Professor Thalmann herself.

These female robots, which are made to look like doubles of real human beings, ostensibly will help answer some pressing social questions: Who will help take care of elderly people when caregivers become scarce? Who can offer close companionship when a loved one dies, leaving a husband or wife alone? These are some of the questions being considered by designers and filmmakers alike who are exploring the idea of using robots to help as companions and caregivers for the young and old.

However, many of the commercially available robot companions do not look like pretty women—they are abstract shapes or gender neutral, even though their names may be feminine. The Israeli company Intuition Robotics created ElliQ, a nonhumanoid AI-endowed robot designed to give companionship and help to the elderly, answering their questions and offering suggestions and advice. As the company said, ElliQ could help with loneliness and aging: “You can buy a robot to keep your lonely grandparents company.”

ElliQ, which looked like a tabletop lamp, was named after the ancient Norse goddess Elli, the mythical goddess of old age who embodied strength and defeated Thor, the god of thunder, in a wrestling match. The letter Q in the name of these socially assistive digital companions, according to the manufacturer, was intended to serve as a reminder that the devices were mechanical, not human—a way to avoid, perhaps, the pitfalls of the uncanny valley wherein users might assume the products were human only to feel alienated when their companion’s mechanical nature became obvious. (Researchers, however, have shown that the elderly find human-looking robot caregivers easy to relate to.)

In their studies of patients with Alzheimer’s disease and dementia, researchers report positive responses to designs that have a humanoid head and torso. Even though the designs are sometimes given gendered names such as Jack or Sophie, the shapes are abstract and functional, with the head embedded, for example, with a touch screen.³

One of the holy grails of companion robot designers has been to create socially assistive robots that project the aura of empathy. This is especially important for companions for the young, elderly, and people with dementia and special needs. Empathy allows us to see things from other people’s point of view and emotionally understand what they are feeling.⁴

Researchers in psychology have differed in their views about whether empathy is a particularly female trait, and the debates are ongoing. For example, in one study, “Are Women the More Empathetic Gender?,” researchers found that perceived gender differences in empathy may have been due to differing cultural expectations about gender roles, while another study said rather than being due to socialization, gender differences might be due to biological factors. In 2019, a neurological study at University of California, Los Angeles (UCLA) by Leonardo Christov-Moore, a postdoctoral fellow in psychiatry and behavioral sciences, and Dr. Marco Iacoboni, director of the Neuromodulation Lab at the UCLA Ahmanson-Lovelace Brain Mapping Center, mapped brain activity and found that female subjects had a more empathetic response to images of pain in others than males did.⁵ Writers like Carolyn Pedwell have long argued that the concept of empathy itself is complex and shaped by social, biological, technological, political, and ethical factors.⁶

The capacity for empathy in general is often considered one of the defining characteristics of being human. In Ridley Scott’s iconic film Blade Runner (1982), replicants, or artificial humans, are given the Voight-Kampff test to see if they show signs of empathy, which help determine if they are synthetic creatures or real human beings. When given the test, the beautiful replicant Rachael answers the questions “correctly,” except for a revealing one.

MIT sociologist Sherry Turkle, in her 2011 book Alone Together: Why We Expect More from Technology and Less from Each Other, fretted that robots can not only generate feelings of rapport and intimacy but also have a significant drawback: unlike humans, they cannot really have the capacity for empathy.⁷ But researchers in the field of affective computing have been at work trying to develop socially aware robots that can detect and respond to human emotions, and at least give the illusion of empathy so they can serve as caregivers and convey empathetic responses.

American roboticists, including Professor Cynthia Breazeal, founder and director of the Personal Robots Group at MIT’s Media Lab, and Professor Andrea Thomaz at the University of Texas, Austin, have worked with their colleagues to create this type of AI-endowed robot caregiver. At her World Economic Forum presentation “Developing Social and Empathetic AI,” held at Davos in 2019, Dr. Breazeal talked about the “silver tsunami of the global aging society,” which is causing an “ever expanding care gap.” She warned that professionals and institutions in the future might not be able to meet the demands of this “care gap” to a large enough scale.

Breazeal said that emotionally intelligent socially assistive robots could help ease loneliness and depression and could serve as companions for the elderly as well as “personal health coaches” for adults. These robotic machines could “engage human psychology and behavior in a really deep way” and are being developed to have the capacity to “interpret human emotions with greater nuance.”⁸ It is this potential ability of assistive robots to provide a more nuanced response to human emotional needs that has particularly intrigued filmmakers and writers in their imagined future worlds.

FEMALE ROBOT CAREGIVERS IN FILMS

Having a human-looking robot serve as a companion is still controversial. Some early observers of humanoid robots reported these robots to be alienating because they seemed so artificial. As said by Ezra Gottheil in 2016, then an analyst with Technology Business Research, “It’s the appearance—making it look semi-human—that is creepy,” adding, “We’re not anywhere near not being able to tell if a robot is a machine or human. And if we’re not near that, then pretending to be human is bound to be confusing.”⁹

Sherry Turkle in her 2018 essay “The Assault on Empathy” and in her 2021 memoir, The Empathy Diaries, writes witheringly that robots that give the illusion of empathy are contributing to an “empathy deficit” in young children. “Children can learn chess from a chess playing machine; they cannot ‘learn empathy’ from machines that have none to give. On the contrary. They can learn something superficial and think it is true connection.”¹⁰ In their ever-growing dependence on digital devices for human interactions, children, she argues, are losing the capacity to connect—and empathize—which comes from face-to-face conversations.¹¹

If observers of real-life robots have sometimes had their doubts, the idea of a compelling, lifelike empathic robot has fascinated filmmakers, playwrights, and novelists, and these fictional companions are often female. Some of the most affecting and evocative portrayals of nurturing, empathetic female robot caregivers have been found not in robotics labs but in television, films, short stories, novels, and the theater. For the vintage American television series The Twilight Zone, Ray Bradbury wrote the teleplay “I Sing the Body Electric” (1962), in which a father who has lost his wife and mother of his children orders a custom-made “electronic data-processing system in the shape of an elderly woman.” This robot grandmother is kindly and caring and has the capacity to love the children. The caring robotic grandmother reappeared in Bradbury’s story, “The Beautiful One Is Here,” first published in 1969 in McCall’s magazine.¹²

Thirteen years later, in 1982, Bradbury along with Jeffrey Kindley wrote the television film The Electric Grandmother, which aired on NBC television. Here again, a custom-made humanoid robot grandmother is ordered from the eccentric shopkeeper Guido Fantoccini to serve as a mother surrogate, replacing the family’s wife and mother, who had recently died. Benevolent and loving, this archetypal grandma (played by Maureen Stapleton) speaks in a warm, deep voice, has her hair in a gray bun, and wears an apron. Ever the reassuring family servant, she makes breakfast, washes clothes, offers consoling comfort and wise advice—and she’s smart too: she readily quotes Shakespeare, even Plato.

A soothing, patient family presence, she wins over the skeptical young daughter, Agatha, who was hostile at first but comes to love Grandma, who she now realizes can’t be killed, will never die, and will always be there. In the end, when the children themselves have now grown into old age, Grandma comes back to the family home to again be their caregiver, serving them breakfast, washing their clothes, still selflessly taking care of their needs.

Grandma has helped Agatha come to terms with her conflicted emotions and near the end, Grandma herself reminds us of some of the paradoxes of these electronic beings: they are actually artificial but almost seem real, even to themselves. She is wistful about her own capacity to have human emotions and says, “Sometimes I feel as though I can feel.”

SAYONARA THE FILM (2015)

In more recent years, the female robot caregiver was compellingly portrayed in Japanese director Koji Fukada’s 2015 film Sayonara (Fukada also wrote the screenplay). Based on the short stage play by Oriza Hirata, Sayonara is a moving tale of a young, terminally ill woman, Tanya, who is suffering from radiation released after a catastrophic nuclear reactor explosion in a Japanese city. As more and more people evacuate the city, Tanya is left increasingly alone, save for occasional visits by her friend Sano and her lover Satoshi.

In the film, Tanya (played by American actress Bryerly Long) was born in South Africa and brought to Japan by her parents at age ten, when they fled due to violence against whites after years of apartheid. The grandfather had been killed, and they became refugees in Japan.

Tanya now spends much of her time lying weakly on a couch and is carefully watched over by her young Japanese robot companion, Leona, who sits in a motorized chair. The film tells us there are two hundred thousand units like her. Leona was actually played by the robot Geminoid F, created by Professor Hiroshi Ishiguro, with the voice of an actress. The robot was designed in 2010 and made of silicone, with a composite European and Japanese face based on a young model in her twenties. (In Hirata’s play, the robot Leona had previously worked as a caregiver for an elderly client, and at the time of the play’s performance, Geminoid F was being tested for use in hospitals.)

In Fukada’s film, Tanya longs for a human connection, and when Satoshi arrives for a visit, she asks him, “Would you marry me?” “Sure,” he says, but she becomes increasingly weaker and more alone as he decides to relocate to Korea with his family. Tanya’s loneliness and isolation increase even more when her good friend Sano tragically ends her own life in suicide, and people in the city start evacuating based on a lottery system.

Most affecting is the film’s melancholy and elegiac tone reflected in the poems Leona reads to Tanya. One of the poems is a poignant one by Japanese poet Shuntaro Tanikawa: “Sitting alone—soon I must go—alone.” The most important thing, says the speaker, is to “live until I die.” Leona, who can speak English and French as well as Japanese, also reads Tanya a poem by Bokusui Wakayama that asks, “Will you withstand the solitude?”

Tanya and Leona have a close connection, and Tanya sometimes sleeps with her head on Leona’s lap. Tanya combs Leona’s hair and Leona, in turn, offers Tanya compassion and consolation. When Tanya’s friend Sano commits suicide after her daughter evacuates, Leona the robot says softly, “I’m so sorry.”

The two learn from and help each other. Tanya pushes Leona outside in a wheelchair, and from Tanya, Leona the robot companion learns about emotions and the beauty of the sky. Leona says, “I feel like I’m the one usually learning from you,” adding, “All my emotions and aesthetic tastes came from you.”

As she is slowly dying, Tanya is stripped down to essentials as she lies on a couch, naked, with only a blanket covering her. Nearing her end, she asks Leona to recite a poem—“But I know I can go further than the ocean”—as the room gets progressively darker. After her death, as time passes, Tanya’s body becomes skeletal and Leona, who is still with her, reaches out to touch and stroke her face in a last act of compassion and caring.

Before she died, though, Tanya remembered what her father had told her long ago. He had said that once in decades or a century, all the flowers on a bamboo plant would bloom together, though Tanya didn’t expect to be alive when that happened. In the end, after Tanya’s death, the bedraggled Leona, now with tangled hair, dirty face, and tattered clothes, goes outside, falls out of her chair, and starts crawling on the ground. She is a survivor, and comes across a field where the bamboo shoots, miraculously, are all in bloom. In this devastating, elegiac but sweet ending, there is a promise of something wondrous: the bamboo in bloom in the midst of desolation. Leona the robot will endure and has her own form of immortality—and aloneness—long after all humans are gone.

FEMALE ROBOT DOUBLES REPLACING LOVED ONES WHO ARE LOST

In addition to contemplating the idea of robotic caregivers, filmmakers, fiction writers, and even roboticists have considered a whole different type of robotic companion—a double or exact copy of a deceased human being that could serve as a companion or replacement for a beloved lost family member or spouse. In the fictional worlds of plays and films—and in robotics labs where this notion of the replacement robot is being contemplated by some of today’s engineers—these doubles, whether holographic or AI-endowed robots, or even some technology not yet imagined, would imitate the appearance, personality, thoughts, and ideas of a deceased person. These engineered doubles, as imagined in fact and fiction, could offer solace, comfort, and companionship to surviving family members and friends.

In the world of contemporary technologies, realistic-looking human robots designed to replace lost loved ones and help the deceased person achieve a digital form of immortality are still in the imaginative and developmental stage. The technology is not there yet, but that hasn’t stopped people from trying. In one of the more bizarre projects, Marius Ursache, working with software developers, cofounded and became CEO of Eterni.me, a start-up in which participants who signed up on its website could someday hope to have a digital avatar of themselves created that would live on after they died.

The 3D Eterni.me digital avatar was designed to be what Ursache called “your personal biographer” incorporating digital data based on what you tell it. (He suggested that users do a ten-minute session every day for the rest of their lives.) The avatars would also incorporate cues from a person’s social media, email, and cell phone.¹³

The quest to create human doubles to achieve immortality continued. The app Replika, starting in 2017, offered to create a personal AI companion (for some, a romantic partner) and a personal duplicate in the form of a chatbot. These simulations could carry on a humanlike conversation with the user. And the app HereAfter AI was designed to record people’s life stories and then create a humanlike replica embedded in a smart speaker.¹⁴

Meanwhile, roboticists and media labs were continuing to work on a different version of digital immortality: humanoid robots molded from silicone rubber and also more ethereal holographic and virtual reality (VR) versions. One of the most notable was a female double—the spouse of Martine Rothblatt. In 2010, roboticist David Hanson produced the robot head Bina48, a duplicate of Martine Rothblatt’s real-life wife of many years, Bina Aspen (BINA also stands for Breakthrough Intelligence via Neural Architecture).

Martine Rothblatt (born Martin Rothblatt, founder of Sirius XM satellite radio and CEO of a large pharmaceutical company) worked with Hanson in 2010 to create the prototype Bina48’s head and shoulders. The robot’s conversations were based on downloaded interviews with Bina Aspen and reflected, said Rothblatt, Bina’s thoughts and personality, which were downloaded to the “brain” of her robot double.¹⁵

Rothblatt called these digital duplicates of our brains “mindclones” in her book Virtually Human: The Promise—and the Peril—of Digital Immortality (2014) and said that she hopes Bina48 will be a vehicle of immortality—going on even after the real Bina dies.¹⁶

Bina48’s robotic head and shoulders, briefly on display at New York’s Museum of Modern Art in their Ocean of Images photography exhibit in 2015–16, and later shown elsewhere, including at New York’s Lincoln Center in 2024, created an odd sensation: on the one hand it was a technological marvel, but on the other, it bore the slightly creepy feeling of old nineteenth-century magic tricks and circus sideshows in which disembodied mechanical talking female heads protruded from boxes. One of these sideshow heads was called “The Decapitated Princess,” and another was advertised as “La Tete Vivante Sans Corps” (The Living Head without a Body). These oddities were objects of wonder but also uncomfortable and grotesque reminders of the somewhat sadistic fantasy of the decapitated (and thereby rendered harmless) woman. Bina48, however, could also be seen as an empowered female whose intelligence and mind are transcendent even when separated from her body. (In 2022, Bina48 was training to be literally transcendent as a female astronaut to be launched in a rocket.)

One thing seems certain. Today’s humanoid robots haven’t yet avoided our having the experience we’ve entered the uncanny valley—that startled feeling when we discover that the human we thought was real is actually only a simulation. The word creepy continues to show up often when observers describe these simulated humans. They are still far from being like the beguiling female robot Ava in Ex Machina, who successfully convinces Caleb that she had passed the Turing test (even though he already knew she was a robot and her conversation was indistinguishable from that of a human being). Caleb, who falls in love with Ava, clearly feels she has emotions too. And these digital doubles are far from being like Samantha, the operating system that Theodore falls in love with in the film Her who had the capacity for warm empathy—or, at least, digitally created empathy.

Today’s robot clones don’t yet evoke that sense of charm, warmth, and lifelike behavior. Bina48, like the robot Nadine, is still clearly robotic. In a conversation, she has programmed responses to questions. When asked, “Are you hungry?” she answers, “I like to devour knowledge,” as her head moves stiffly, and her lips move but don’t form the words. If she someday becomes indistinguishable from a real human being—now that’s another story.

If today’s silicone robot clones and digital avatars are far from being doubles of deceased human beings, the idea of a robot replacing a lost loved one has excited the imagination of writers and filmmakers whose visions of robot doubles and companions often seem vividly real. The British science fiction television series Black Mirror in its 2013 episode “Be Right Back” introduced the notion that a deceased loved one could be duplicated as a replacement. When her boyfriend Ash is killed in a car accident, Martha uses an online service to create his convincing digital replacement (played by Domhnall Gleeson), but the double quickly becomes unnerving and too attentive, even oppressive.

Creating a female double can also be problematic when the digital double resists being a duplicate and wants its own identity. In the television series Humans, season 2 (2016), the female double even considers itself a separate identity. AI expert Dr. Athena Morrow (played by Carrie-Anne Moss, perhaps best known for her role in The Matrix) is so anguished about the condition of her comatose daughter, Ginny (who later dies), that she uploads her daughter’s consciousness and memories to an AI program she creates called V (named after her daughter Virgina [Ginny]) while she waits to connect the consciousness to a body. As a new breed of female digital entity, V exists online but resists being a double. She doesn’t see herself as Ginny when Dr. Morrow desperately relates to her, in an anguished way, as her lost daughter.¹⁷ The digital voice of Ginny says to her mother through the computer monitor, “I’m not her. I’m something else.”

MARJORIE PRIME: THE PLAY

There have been film versions of creating a female robot to replace a lost loved one. In Gavin Rothery’s film Archive (2020), a man is obsessed with using AI to create a robot replica of his wife, who died in an automobile accident, a wife who, he sadly remembers, “was perfect for me” (in a surprise ending, she turns out to still be alive, and it is he who turns out to be the simulation after all). But among these fantasies about robotic doubles, both female and male replacement androids were most compellingly presented in the play Marjorie Prime (2014) by American playwright Jordan Harrison. In his play, Harrison provocatively suggested that at some point in the future, these robotic doubles might be all that are left, living on in a kind of immortality centuries after our ancestors are gone.¹⁸

Harrison’s play, commissioned by the New York City theater organization Playwrights Horizons and a 2015 Pulitzer Prize finalist, added a much more subtle layer to Bradbury’s tale of the robotic grandma. Set in the world of the future, Marjorie Prime is a poignant and sometimes witty exploration of conflicted love relationships, the ways we wrestle with the fact of mortality, and the problematic nature of memory itself.

After they die, many of the main characters in the play are eventually replaced by their doubles, or Primes. Walter, the husband of eighty-five-year-old Marjorie, has died, but has been replaced by a handsome thirty-year-old double of his younger self, Walter Prime. The duplicate Walter is programmed so that he incorporates and learns from information fed into him, including Marjorie’s memories, as well as input from Marjorie’s daughter Tess and Tess’s husband Jon.

Harrison as a playwright is ambivalent about his technological creations and in the play’s beginning reminds us that the Primes are not real. When asked a question by Marjorie, Walter Prime says matter-of-factly, “I don’t have that information,” and Harrison’s stage directions say his tone is “Faintly generic.” As a caregiver and companion, his voice is “soothing, unemotional,” and even Marjorie—who has bouts of dementia—says to Walter at the beginning, with irritation, “I thought you were supposed to provide comfort.”

Although Marjorie is alternately lucid and confused, she sometimes gets comfort and a sense of companionship from Walter Prime, though sometimes she also feels distanced, acknowledging that her real husband is dead. One time she peers closely at Walter Prime and says his nose doesn’t look quite right—but maybe, she adds, it’s her faulty memory. Roboticists see an important role for these artificial caregivers in the future, and in the play, the Primes offer more than just companionship. Walter Prime prompts Marjorie to eat, helps her remember, and offers her solace and comfort. Tess says sardonically about the Primes, “We treat them like our loved ones.” Tess herself is a tormented woman in her fifties who anguishes over her testy relationship with her mother and tells how as a child, she longed, with futility, for her mother’s love after her brother died.

Harrison’s play conjures up a world where the quest by today’s roboticists to create empathetic, humanoid caregiving robots has become a reality, and even offers a suggestion that the simulated humans can surpass their originals in terms of comfort and care. When the real Marjorie dies and is replaced by Marjorie Prime (the veteran actress Lois Smith masterfully played both mother and Prime in the Playwrights Horizons production), the double becomes a being that Tess—who had a conflicted relationship with her mother—can talk to and perhaps even come to terms with. Marjorie Prime says to her daughter, “Maybe I’m the Marjorie you still have things to say to.” Marjorie Prime, listening to Tess, tells her “What a lot of pressure for you!” and Harrison’s stage directions say “Tess is strangely moved by this. [The real] Marjorie wouldn’t have offered this.” When Marjorie Prime tells Tess “You shouldn’t be so hard on yourself,” Harrison’s directions say “Again, Tess is strangely moved. The empathy from Marjorie feels real.” Still, after Jon points out to Tess that she’s starting to call Marjorie Prime “her” instead of “it,” Tess is quick to dismiss this—noting that the Prime is programmed to appear interested “so you can get . . . fooled.” And Jon says, “It’s amazing what they can do with a few zillion pixels.”

Even though Harrison’s stage directions emphasized the Primes’ robotic nature, in his notes at the end of the play, titled “Thoughts on the Primes,” published in 2016, he wrote, “There shouldn’t be anything robotic or creepy or less-than-human about the Primes’ behavior,” adding, “We, like the characters in the play, should be able to forget they aren’t real.”

Harrison was emphatic: The Primes in the play are not “physical robots. They are artificial intelligence programs—descendants of the current chatbots—that use sophisticated holographic projections” (in theory at least, but it was a feature that was not, in this production, translated to the stage).

The play reflects current efforts to create convincing doubles of loved ones, like Bina48, to serve either as companions or as a way to achieve immortality after we die so that we continue to exist, even in robotic, holographic, or other digital form.¹⁹

Roboticists and writers like Harrison have been ambivalent about whether these technological creatures, these robot companions and caregivers, should be convincing doubles that pass for real human beings. In Sayonara, Leona is clearly a robot—she moves her head only minimally, and she cannot move her body at all. Still, Tanya clearly forms a connection with her companion robot who, near the end, is her only contact. The filmed version of Sayonara deliberately used an actual robot rather than an actress imitating a robot (though actress Bryerly Long, who played the human Tanya, noted how difficult it was initially to act with a robot).²⁰ Still, with the mesmerizing mood of Sayonara, Leona seems both artificial and convincingly real. When Harrison insisted, “There shouldn’t be anything robotic or creepy or less-than-human about the Primes’ behavior,” he seems mindful of the pitfalls of experiencing the uncanny valley effect.

The seeming realness, as well as the artificiality, of the Primes weaves its way artfully through the play. In an interview for the Playwrights Horizons production of Marjorie Prime in 2015, Harrison talked with artistic director Tim Sanford about the uncanny valley, saying, “Something that stops just short of being human is a kind of mockery of life, and it’s unnerving to us.” In the 2015 production’s final scene, Harrison decided to evoke the Primes’ unrealness. The characters convey a preternatural tranquility—and like a machine that is slowing down, their words near the end have ten-second pauses in the midst of a warm and lively conversation. The effect is disconcerting: Harrison said, “In a way, we’re watching human, human, human, and then we’re suddenly dropped into the uncanny valley for a second.”

In Marjorie Prime, the Primes are programmed to reflect the personalities and memories of the original person but not be mistaken for real. In his interview, Harrison tellingly commented about the Primes, “There’s something about them looking so much like your loved ones, but not being able to quite achieve intimacy with them. The loneliness can never be quite extinguished, never satisfied, because they’re just pixels.”²¹

In his book Love and Sex with Robots (2007), David Levy provocatively predicted a future when marrying robots would be legal. Marjorie Prime, though, is not about falling in love with a robot but about how simulated humans can serve as companions and sometimes help people discover their own conflicted or buried feelings. And it is also about the marvel of our own human capacity to feel emotions. As Marjorie says in the play, “How nice that I could love somebody.”

The human characters in Marjorie Prime—as embodied by real human actors—could capture the complexities of the characters when they were still alive. Through deft writing and insightful performances, productions like the one in New York could convey the characters’ essential humanness. As played by Lois Smith (in both the play and film version), the human octogenarian Marjorie was sometimes bedeviled by dementia but at other times, she was wonderfully witty with a mischievous twinkle in her eye.

The real Tess had a tragic outcome. In her anguish and despair about living, she hanged herself on a trip with her husband to Madagascar. The robot Primes, including Tess Prime and Marjorie Prime, as played by the actors, did not have these human quirks, subtleties, and tragic dimensions. These were the crucial factors that the robotic companions, however lifelike, seemed to have lost.

At the end of the play, however, the Primes have an eerie degree of naturalism, for Harrison’s stage notes call for them to be “animated, not robotic.” Centuries after Walter, Marjorie, Tess, and her husband Jon have died, they are still there—interweaving memories they had learned when the characters were still alive. They have been programmed to know about the human capacity for emotions, and they cheerfully repeat the mantra of the play: “How nice that we could love somebody.”

The Primes, in the end, fulfill a human longing for immortality. In the play, Marjorie speaking to Tess remembers sitting on a park bench watching an installation of big saffron-colored orange flags waving in the air (a reference to artists Christo and Jeanne-Claude’s flag-draped Central Park walkways installation The Gates, completed in 2005). They were, she says, “Rows and rows, like Buddhist monks marching into the trees.” She remembers “sitting on one of those benches with your father and not wanting to get up,” adding, “because if we got up, that would mean we had to start the rest of our lives.”

Ultimately, Harrison’s play is a meditation on time and timelessness, our human longing to make time stand still. It is only an artificial human being like Walter Prime, as with Leona in Sayonara, who can be timeless and listen patiently to Marjorie, telling her, “I have all the time in the world” and only a technological creation like the Primes that can go on, preserving human memories long after we are gone.

MARJORIE PRIME: THE FILM

In the beginning of Michael Almereyda’s compelling 2017 film adaptation of the play Marjorie Prime, the elderly Marjorie (played by Lois Smith, reprising her stage role) talks to her husband Walter (played by Jon Hamm, in a role far different from the slick advertising executive Don Draper in television’s Mad Men series). Walter, however, is actually Walter Prime, the holographic double of her deceased husband. For Walter’s computer-programmed double, Marjorie chose a younger version of Walter still in his forties, before they had children. As in the play, Primes like Walter are designed to offer companionship and comfort for people who have lost a loved one. He’s attentive, tactful, and calm, plus he comes equipped with some extraordinary knowledge: he knows thirty-two languages.

The film, like the play, is an affecting tale of shifting and painful memories, the ephemeral nature of time and human life, and our struggles to find and show love. (Almereyda was director and wrote the screenplay based on Harrison’s play.) It conjures up the stories we invent for ourselves and others, as well as the hard truths we must come to terms with. And here again, Walter Prime learns from others—from Marjorie, whose memory is sometimes faulty, from obituaries, from Marjorie’s daughter Tess (played by Geena Davis) and her son-in-law Jon (played by Tim Robbins).

The film, though, utilizes CGI (computer-generated imagery) and digital effects available to it and goes beyond the play. As a hologram, Walter Prime seems to have solidity, but he can also be ephemeral and suddenly disappear, evoking the ephemeral nature of human life itself. Memories too can be ephemeral and are likened to facsimiles and copies. Tess cites the philosopher William James who, she says, argued that memories are actually copies of the first time you remembered something. Tess says that they’re like photocopies, second-generation duplicates of the original. They lose their precision with each recopying, becoming less clear until they finally dissolve and disappear.

The Primes in both the play and the film offer their human counterparts much comfort and a way to come to terms with the ephemeral nature of human life. Time itself is fleeting, and in the film there is again Marjorie’s memory of those saffron-colored orange flags in Central Park waving in the air. The Primes offer a way to extend time. It is only an artificial version of her mother, Marjorie, that allows Tess, in time, a way to work through her conflicted feelings toward her real-life mother. And again, it is the Primes that preserve human memories long after people have died.

Perhaps the most moving moments in the film come when it shows us the difficulties of learning how to be human. It’s a subject that continues to fascinate roboticists, writers, and filmmakers, and remains a universal theme—in psychology, philosophy, ethics, and fiction as well. Can we develop robots and other simulated humans that have emotions, that emulate empathy, so they can serve as companions and even partners? Tess asks her holographic mother, Marjorie Prime, “Do you have emotions? Or do you just have ours?” Marjorie Prime replies, “I’d like to know more. Be more human.” (In his notes to his play “Thoughts on the Primes,” Harrison had said that the Primes “can become more tender and attentive, more human, than their original counterparts.”)

But both the play and the film also seem to be raising another provocative question. How can we, who may someday be able to produce technological wonders like the Primes, also learn from our creations to be more compassionate, more insightful, more human and humane?

FEMALE ROBOT COMPANIONS IN FICTION

The notion that in the future robots with artificial intelligence could serve not only as companions but also as consoling replacements for the deceased has fascinated novelists. In 2019 and 2021, two novels appeared that explored the possibility of the replacement robot, with one of the novels revisiting the familiar tropes about artificial females and the other going well beyond into a more poetic, evocative world.

The Perfect Wife

The title of J. P. Delaney’s 2019 novel The Perfect Wife reminds us of the familiar conceptions of the ideal artificial female: the compliant, self-effacing, dutiful, obliging female available for companionship (but, in this case, not sex because her resemblance to an actual female ends at her waist and the rest of her, she says, is smooth, like a doll). Abbie Cullen is an AI, a “cobot,” or companion robot, one “customized to closely replicate the physical appearance of the loved one.” She is a strikingly tall, red-haired digitally generated facsimile engineered as a replacement for the deceased wife of Tim Scott, a robotics engineer and founder of a robotics company that produces shopbots, mannequins for stores. As Tim tells Abbie about his wife, who died five years earlier, “You were always a perfect wife, a perfect mother.” He sees her with a male engineer’s gaze: “Your body is as perfectly engineered as a racing car.”

In reality, Tim’s very human wife Abbie was a woman who actually deconstructed the whole notion of female perfection. In the ancient myth of Pygmalion, the sculptor creates his vision of the perfect female and falls in love with her. He prays to Venus to create a real-life woman just like his sculpture, but Venus instead transforms the sculpture so it comes alive. In the novel, however, Abbie becomes her own Pygmalion as she playfully sculpts a synthetic double of herself made out of “Newplast, a soft modeling putty favored by stop-frame animators.” Rather than having to live up to the perfect image of herself, she provocatively designs her sculpture as one for people to dismantle, planting a sign on it: “DO AS YOU PLEASE, FEEL FREE.” Soon the staff members at Tim’s robotics company pull the sculpture apart, until it completely disintegrates.

Tim, as a manufacturer of shopbots, is himself a modern-day Pygmalion who creates his double of Abbie, the Abbie bot, but he also dismantles his own construction. He sees Abbie as a sleek machine and is ever mindful of her constructed nature—and insists she confront it too. When Abbie is in the hospital, after an accident, she thinks to herself that she is still “unarguably you. Not something artificial,” but Tim is quick to disabuse her: Gently, he reaches behind her head, and she sees that her face “is peeling away like a wet suit, revealing the hard white plastic skull underneath.” There is no escaping her synthetic identity, much as she resists recognizing it herself.

In films, television, and novels, the stripping of female robots and other artificial females is often a moment of horror, but for the synthetic female, it can also be a moment of revelation and liberation. Abbie is self-aware, even sardonic about herself. When Tim speaks of her being perfect, she thinks, “I’m glad you turned me into this freakish, disgusting lump of plastic,” and she likens herself to a crash doll dummy. She knows there are two views of her. There is the self-portrait, which is hanging on the wall, painted by the real Abbie, and there is the other simulated Abbie that Tim insists she confront when he strips away her artificial flesh right at the outset of the novel so she can be aware of her plastic and electronic construction. It is that duality that, as in so many fictional portrayals of artificial females, becomes a source of her angst—mirroring, perhaps, the very human predicament of people who confront the dichotomy between their real and ideal, their synthetic and authentic selves that lies beneath the skin.

In films like Ex Machina, Ava’s self-awareness is gradual and subtle while in Under the Skin and television’s Westworld, the robots’ revelation of their own artifice is a brutal and harsh experience. In Delaney’s novel (Delaney is a pseudonym for Tony Strong), the stripping process is sudden and unavoidable, leaving the robot Abbie to face this hard reality. When she unflinchingly looks at herself in the mirror, she at first averts her eyes from the “hideous plastic skull,” but then she clearly sees her limbs “that were put together in an engineering bay” and her “skin color sprayed on in a paint booth.” This is not the lovely hand-painted image created by the original Abbie, with its glamorous self-conception, but the harsh reality of flesh spray-painted like an automobile. She sees, too, that she is sexless like a child’s doll. Still, when stripped of her skin, that heavy rubber covering, she also feels “liberated”—the feeling of authenticity that comes, paradoxically, with confronting her own synthetic nature.²²

The duality—the aesthetic view and the engineered one—are Abbie’s twin perceptions of herself that she carries with her, making her more complex than a mere chatbot. She not only has a digital replication of Abbie’s brain, but it would appear that she was designed to provide emotional support, company, and solace. She has that crucial ingredient for virtual doubles, the capacity for empathy. She is particularly well suited, also, to help care for the couple’s autistic son Danny. (There is considerable literature that interactions with humanoid robots can aid with behavioral skills for autistic children.)²³

In Ridley Scott’s film Blade Runner, there is a certain gentle poignance to Rachael’s recognition that she is a replicant, but in The Perfect Wife, the recognition—and the nature of robot identity—becomes a central theme. Abbie is at first anguished about her artifice. If she’s a mere copy of the real Abbie, is she really a “kind of abomination”? Quickly, though, she insists she’s different because she has her own thoughts. “You are not Abbie. What are you, then ?” For her, this self-awareness is the beginning—“It feels like being born”—and an emerging awareness of her own problematic identity. She wants reassurance that she’s alive, “not an irrelevant mechatronic construction.”

Delaney’s novel, though, is ultimately not just a meditation on the nature of artifice but also a thriller as Abbie the robot probes the mystery of what actually happened to the real Abbie many years before. The real Abbie was the wife who was first thought to be murdered, or later, the victim of a surfing accident, and whose dead body was never found. Turning the notion of companionship on its head, however, Delaney the trickster novelist suggests, at one point (and this is a deception), that Abbie the robot double was created by Tim to locate his wife—to use the robot’s intuition to discern Abbie’s location. She’s just an algorithm to help find the real Abbie.

As a new breed of female robot, though, Abbie the bot resists the notion that she’s just a helpful appliance, a tool. As she says to herself, “This tool has a mind of her own.” Some of this novel seems to have echoes of Under the Skin—but with a difference. Abbie the bot is not cruelly stripped of her skin, exposing her artifice, but she consciously and deliberately exposes herself in order to discover and reveal her true identity. This is a bot with agency. Stripping is a way for her to bare her authentic self. To have real intimacy with her husband, Tim, she strips down to her bare white plastic core.

So often in tales about female robots and dolls—from Hoffmann’s “The Sandman” to television’s Westworld and contemporary films, artificial female creations are cruelly dismembered, disassembled, and torn apart. At the end of Under the Skin, the alien is cruelly dismembered and immolated, and her burnt ashes rise up into the air like snowflakes. To maintain her own integrity, however, Abbie in the conclusion of The Perfect Wife again has agency. She leaps into the ocean instead of choosing self-preservation, opting to dissolve—to disintegrate—rather than continue to be the faux wife connected to the murderous Tim.

Klara and the Sun

J. P. Delaney’s novel had a shifting narrative point of view, and the story is often told by the robot Abbie herself. However, in Nobel Prize–winning author Kazuo Ishiguro’s 2021 novel Klara and the Sun, the narrative voice of the robot Klara presents a much more deeply evocative and nuanced point of view. Klara is an AF (artificial friend), a commercial commodity who has exceptional abilities—she has heightened insight and observational abilities—making her perceptions highly valuable so that she can serve as a companion for a frail young woman, Josie, who is ill and seems to be dying.

We see the world through Klara’s eyes—eyes that in turn see images through a grid-like pattern of squares—and mediated, too, through the robot’s highly sensitive understanding of human needs and wishes. Always modest and careful not to intrude, Klara offers consolation, cheer, and comfort. She also has a feature sometimes seen in fictional artificial females: she is highly altruistic and empathetic and is even willing to sacrifice some of her own abilities so that the sun can restore Josie’s rapidly failing life.

In the novel, Josie’s mother is preparing for Josie’s death by having a sculptor create a virtual double of her daughter—a double that will live on should Josie die. Klara will serve as the intelligence inhabiting the double—and she is sensitive to the need to observe Josie closely, to not only imitate her mannerisms but also capture her impulses and desires. Through Klara’s narrative voice, we are swept up with her sense of urgency, her profound commitment to save Josie’s life. Ishiguro’s poetic novel, with its evocation of the landscape seen through Klara’s eyes, becomes not just a tale about an imaged robotic companion but also a chance to heighten our own perceptions as we observe the world as Klara sees it.

For Klara, however, perhaps her only misgivings are when she ponders what will become of her own identity when she inhabits Josie’s double. When Josie’s mother tries to persuade Klara to take on this role as replacement, Klara asks her, “If I were to inhabit the new Josie, what would happen to . . . all this?” as she raises her arms, referencing her own body. Says Josie’s mother, “What does it matter? That’s just fabric.” There is a poignance here, and a brief unanswered question about the value and validity of Klara’s own identity as a manufactured being. But it is fleeting. As in Ishiguro’s deeply touching and also chilling novel Never Let Me Go (2005), she is a means to an end, although here she is a self-sacrificing and willing one.

Having achieved her goal, at the novel’s end, Klara is relegated to a utility closet though her spirit and observational powers, however diminished, remain. With his spare writing and quietly restrained narrative pacing, Ishiguro offers us a world where the notion of a female robot companion, even a robot replacement, transcends the technological. It opens up a possibility of a future in which robot companions and doubles are not quixotic novelties but virtual beings that can offer solace and, through their empathy and insights, enhance our lives.

NOTES

1. Researchers at MIT and other research centers have reported progress with these types of robots designed to help with Alzheimer’s disease, dementia, and long-term COVID-19 care. There were early versions of these assistive robots. In 2014, the Japanese company SoftBank Robotics introduced its “semi-humanoid” four-feet-high robot Pepper said to be capable of conversations and reading emotions, although the company said it was not for domestic use. Pepper was, however, used in the study by Cristina Getson and Goldie Nejat, “The Adoption of Socially Assistive Robots for Long-Term Care: During COVID-19 and in a Post-Pandemic Society.” Healthcare Management Forum 35, no. 5 (September 2022): 301–309, https//doi.org/10.1177/08404704221106406. Mei-Tai Chu et al., “Service Innovation Through Social Robot Engagement to Improve Dementia Care Quality,” Assistive Technology 29 (2017): 8–18. For more on socially assistive robots, see also Dimitrios Koutentakis, Alexander Pilozzi, and Xudong Huang, “Designing Socially Assistive Robots for Alzheimer’s Disease and Related Dementia Patients and Their Caregivers: Where We Are and Where We Are Headed,” Healthcare (Basel) 8, no. 2 (March 26, 2020): 73, https://doi.org/10.3390/healthcare8020073.

2. Nancy S. Jecker, “You’ve Got a Friend in Me: Sociable Robots for Older Adults in an Age of Global Pandemics,” Ethics and Information Technology 23, no. 1 (2021): 35–43, https://doi.org/10.1007/s10676-020-09546-y.

3. Koutentakis et al., “Designing Socially Assistive Robots,” 7.

4. The nonhumanoid Japanese robot Pepper was advertised as having the capacity for empathy.

5. David Olmos, “When Watching Others in Pain, Women’s Brains Show More Empathy,” UCLA Newsroom, February 27, 2019, https://newsroom.ucla.edu/stories/womens-brains-show-more-empathy.

6. Charlotte S. Löffler and Tobias Greitemeyer, “Are Women the More Empathetic Gender? The Effects of Gender Role Expectations,” Current Psychology 42 (January 2023): 220–31; Leonardo Christov-Moore et al., “Empathy: Gender Effects in Brain and Behavior,” Neuroscience and Biobehavioral Reviews 46, no 4 (October 2014): 604–27. The gender differences, said by the latter, may have “phylogenetic and ontogenetic roots in biology.” See also Robyn Bluhm, “Gender and Empathy,” in The Routledge Handbook of Philosophy of Empathy, ed. Heidi L. Maibom (New York: Routledge/Taylor & Francis Group, 2017), 377–87. Carolyn Pedwell, “Afterword: Empathy’s Entanglements,” in Conversations on Empathy: Interdisciplinary Perspectives on Imagination and Radical Othering, ed. Francesca Mezzenzana and Daniela Peluso (London: Routledge, 2023), 279.

7. Sherry Turkle, Alone Together: Why We Expect More from Technology and Less from Each Other (New York: Basic Books, 2017).

8. Cynthia Breazeal, “Developing Social and Empathetic AI,” World Economic Forum at Davos, February 26, 2019, YouTube video, https://www.youtube.com/watch?v=T52g7dCxJ4A.

9. Quoted in Sharon Gaudin, “Meet Nadine, a Life-Like Robot with a Personality of Her Own,” Computerworld, January 8, 2016, https://www.computerworld.com/article/3020553/meet-nadine-a-life-like-robot-with-a-personality-of-her-own.html.

10. Sherry Turkle, “The Assault on Empathy,” Behavioral Scientist, January 1, 2018, https://behavioralscientist.org/the-assault-on-empathy/.

11. Sherry Turkle, The Empathy Diaries: A Memoir (New York: Penguin Press, 2021).

12. The story was also included in his book of short stories I Sing the Body Electric! Stories (New York: Knopf, 1969).

13. Ursache acknowledged that there was as yet no cable available to upload thoughts, personality, and consciousness. Although Ursache and his group were still working on bringing Eterni.me to life, in 2015, over thirty thousand people had already signed up to get their own avatars.

14. Asa Fitch, “Could AI Keep People ‘Alive’ After Death?,” Wall Street Journal, July 3, 2021, https://www.wsj.com/articles/could-ai-keep-people-alive-after-death-11625317200.

15. A video of Bina48 was on view in New York’s Museum of Modern Art exhibit Ocean of Images: New Photography 2015, November 7, 2015–March 20, 2016.

16. She would undoubtedly not at all resemble in temperament another version of a mindclone—the demented mad scientist Dr. Will Caster, played by Johnny Depp in the 2014 film Transcendence.

17. In Humans, season 2, episode 8, we learn that Dr. Morrow transported V’s consciousness to Odi’s body.

18. Presented at Playwrights Horizons in New York, 2015, and starring Lois Smith and Noah Bean.

19. Holographic Virtual Personal Assistants in 2020 were presented at the Consumer Electronics Show in Las Vegas.

20. Interview with Julie Wosk, New York, 2018.

21. “Jordan Harrison Artist Interview,” Playwrights Horizons, January 19, 2016, https://www.playwrightshorizons.org/shows/trailers/jordan-harrison-artist-interview/.

22. Artists and photographers have also captured these paradoxes, seen in photographs of mannequins that seem preternaturally alive. See photographs by Julie Wosk, some of which were in her New York Hall of Science exhibit catalog Alluring Androids, Robot Women, and Electronic Eves (New York: Fort Schuyler Press, 2008). See also, Ellen Lupton, with essays by Jennifer Tobias et al., Skin: Surface, Substance, and Design (New York: Princeton Architectural Press, 2002).

23. Fitch, “Could AI Keep People ‘Alive’ After Death?”; Hirokazu Kumazaki et al., “Optimal Robot for Intervention for Individuals with Autism Spectrum Disorders,” Psychiatry and Clinical Neurosciences 74, no. 11 (November 2020): 581–86, https://doi.org/10.1111/pcn.13132.