Directed by Anne Kauffman, the play features “Primes,” or holographic simulations of the dead intended for therapeutic use by the living. June Squibb—who, at 96, is making history as the oldest performer to open a Broadway show—astonishes as Marjorie, an impish 85-year-old with dementia using a much younger version of her husband Walter (an uncanny yet tender Christopher Lowell) to regain and retain her memory. Marjorie’s daughter Tess (the incredible Cynthia Nixon) is skeptical and fearful of the technology, whereas Tess’s husband Jon, played by a standout Danny Burstein, is a fan—until an on-the-nose change of heart in the penultimate scene. 

Marjorie Prime’s central flaw is that it favors concepts over dramatic depth. The characters are well-acted but underdeveloped, and almost all they do is talk; the biggest event may be Marjorie urinating herself. Yet, despite its slow pace and formulaic structure, Marjorie Prime is intelligent and poignant. 

Marjorie’s memories are embellished and sanitized for her comfort and convenience. The fallibility of memory is hardly a novel concept, but the Primes enable this reconstructive process and also become a stand-in for genuine connection in the wake of grief, preventing the family from confronting painful realities and repairing their relationships.  

By the time the truth fully surfaces in the unsettling final scene, which makes adroit use of a stage turntable (props to scenic designer Lee Jellinek), there are no humans left to heal. When storytelling is delegated to AI, truth becomes archival rather than relational; however, truth must be witnessed between living people in order to be ethically and therapeutically meaningful.  

Playwright Jordan Harrison’s Primes, like flesh-and-blood clinicians, absorb and co-construct patients’ accounts of self, yet they are disembodied, unfeeling, and ultimately unable to act with compassion, turning dynamic stories into datasets.   

Marjorie Prime, through Feb. 15 at the Helen Hayes Theater in New York; 2st.com/shows/marjorie-prime.  

“Much of healthcare happens in interpersonal moments,” write Maura Spiegel and Danielle Spencer in the first chapter of The Principles and Practice of Narrative Medicine—and machines are good at many things, but participating in a truly interpersonal moment is likely not one of them. Several studies have suggested that models perform worse for underrepresented groups because they are trained on datasets that lack racial, cultural, and linguistic diversity. Additionally, AI may miss subtle emotional cues and fail to interpret tone, context, and metaphors, which, one bioethicist [1] predicts, could “fundamentally alter” how trust is practiced in healthcare. Others [2] have underscored that “AI should be viewed not as a replacement for the physician, but as a partner in delivering empathetic, patient-centered care.” 

However, AI is not wholly bad. For example, a recent systematic review [3] found that applying natural language processing (NLP) to unstructured text in electronic health records (EHRs) can detect signs of cognitive impairment. Some [4] have found solace in text-based simulations with lost loved ones. And perhaps technology should be viewed as a vehicle for strengthening partnerships between clinicians and patients. Designed by Gabriela Gomes, the video game Healing Spaces aims to help those with neurodegenerative diseases connect with their caregivers. It is a multisensory experience: an app with beach and forest scenes, and a box with aromatherapy that smells like pine trees. Healing Spaces may evoke memories or even create new ones between caregiver and patient, unlike the Primes’ hollow curation. 

Healing Spaces also includes sunscreen-scented lotion that caregivers can use to massage the hands of those in their care. Needless to say, holograms and lotion don’t pair well. “You can’t touch a hologram. So there’s something about them looking so much like your loved ones, but not being able to quite achieve intimacy with them,” said Harrison during Marjorie Prime’s Off-Broadway run about a decade ago. “The loneliness can never be quite extinguished, never satisfied, because they’re just pixels.”  

[1] Kerasidou, Angeliki. “Artificial Intelligence and the Ongoing Need for Empathy,  Compassion and Trust in Healthcare.” Bulletin of the World Health Organization, vol. 98, no. 4, 2020, pp. 245-250. pmc.ncbi.nlm.nih.gov/articles/PMC7133472/. 

[2] Ghenimi, Nadirah, et al. “Integrating AI with Narrative-Based Medicine: Enhancing Patient-Centered Care in Primary Practice.” Perspectives in Primary Care, 5 Dec. 2024, info.primarycare.hms.harvard.edu/perspectives/articles/integrating-ai-with-narrative-based-medicine.  

[3] Shankar, Ravi et al. “Natural Language Processing of Electronic Health Records for Early Detection of Cognitive Decline: A Systematic Review.”npj Digital Medicine, vol. 8, no. 1, 2025, p. 133. pubmed.ncbi.nlm.nih.gov/40025194/. 

[4] Fagone, Jason. “The Jessica Simulation: Love and Loss in the Age of A.I.” The San Francisco Chronicle, 23 July 2021, sfchronicle.com/projects/2021/jessica-simulation-artificial-intelligence/. 

Web image from 2nd Street Theater.

I am glad to report that Playground is a success on this count. In this most recent novel, Powers turns to the ocean as a backdrop for his story. He weaves together the narratives of three main characters. Todd Keane, who is the son of a successful businessman, grows up privileged in upper-crust Chicago. As a high school student in a top-tier private school in the 1970s, he is captivated by the relatively crude computers of the time, mounts the technology wave and feverishly rides it to professional success as an adult. Evie Beaulieu is literally and figuratively thrown into the water at a young age by her father, a deep sea explorer, and becomes obsessed with everything watery. She is more comfortable swimming deep beneath the ocean surface than walking on land. With the support of a gentle and self-effacing husband, she becomes a world famous oceanographer who achieves renown because of her ability to convey the magic and mystery of the ocean equally well to her scientific colleagues and young adult readers, echoes of Richard Powers himself. Finally, Ina Aroita grew up on the Pacific island, Makatea, but moved to the United States to pursue studies in the creative arts and becomes a sculptor. In college, she meets Todd and through him she is introduced to an important fourth character, Rafi Young, who pulls together the Keane-Aroita story lines.

Rafi grew up in the Black neighborhoods of South Side Chicago. He is a voracious reader from the first day of school and, defying the odds,  wins a scholarship to attend the same high school as Todd Keane. They meet over intense games of chess, advance to the ancient addictive game of Go, and develop a powerful but vulnerable friendship. They both choose to go to college at the University of Illinois, and Ina is drawn into their orbit.  Powers pulls all of these characters into a tight web that centers on the ambitious plan of an American company  to build a self-sustaining city that will launched from Makatea and will be submerged under the surface of the water. The residents of the island are being asked to weigh the pros and cons of the “seasteading” initiative and decide whether to approve it. The rationale offered by the proponents is that life in such a unique environment will enable the occupants of the submerged city to live in a realm outside the jurisdiction of any company and be free to pursue their intellectual pursuits and dreams about expanding the use of artificial intelligence, free of government interference and restrictions.

Extreme libertarianism meets AI wonderland.

Powers populates his literary world with characters we come to know well and grow attached to. It is a strength of Playground that they are believable and not simply spokespeople for a point of view. Powers explores many themes in this book. Games, those that people play casually for leisure, like chess, and those that they play intensely for success in life figure prominently in the novel. Todd and Rafi’s obsession with the ancient board game of Go, with its infinite number of possible outcomes branching out from a simple 19×19 linear grid, becomes an incubator for their ideas about computing capacity. Go provides an alluring framework to explore human behavior and model it for use by programmers of life-like computer surrogates. The game provides the initial inspiration for Todd to design powerful AI systems that can mimic humans and beat them at their own games. There is an appreciation of the growing potential for unpredictable outcomes as computer systems become more complex, machine learning diversifies, and neural networks become denser. The example that most impresses Keane (and Powers) involves the game of AlphaGo, an actual computer program, which was designed to play Go at championship level. Todd recounts his feelings of awe when the program  made an inexplicable move in a game against Ke Jie, the number one ranked player in the world at the time, that sealed the computer’s victory in a three-game match. This was something that was thought to be beyond the power of any computer (this same episode features prominently in the book Maniac by Benjamin Labatut).

In this novel, the inventiveness of nature and the unpredictability of human interactions far surpasses anything that AI accomplishes. The complexity of the natural world, its wondrous diversity, its unexpected vitality, is captured in incandescent prose as Powers describes the glowing colors of the deep-sea creatures. Can and will computers achieve a sense of stewardship and legacy? Will they develop an attachment to the environment around them and struggle to sustain it and pass it on whole to the next generation of computers in the same way that links people across and with future generations?

The novel reaches a climax and the ending involves an unexpected plot twist. The book is too much of a fun read for me to reveal it and spoil it for you. One hint — keep in mind that Todd Keane is an extraordinarily complex person. That said, what is striking is how marginally AI actually influences the actual outcome of the novel. It is more of a gadget play than an active player in the narrative. It is as if Powers himself is unsure how AI will impact those of us alive today and the generations that will come after us. If this is true for Powers, then it suggests that all of us should stay modest in our predictions of the future of AI. Reading Playground, I think he would remind us to keep in mind that however things turn out with AI, it will probably work out best if we remember the grandeur of nature and the human capacity to care deeply for one another and our environment, qualities that Powers describes with an expansive intelligence and poetic beauty.

Playground
Richard Powers
W. W. Norton & Co.2024.
381 pp (paperback)

Web image by Medhum.org

While only in nascent forms, organizing a method for understanding artificial intelligence (AI) as it’s currently structured for broad application is important but daunting. Shannon Vallor, a philosopher and ethicist, suggests using a mirror as a metaphor in her book, The AI Mirror. The book stems from her work as both an academic philosopher and as an AI ethicist at Google.  

Vallor describes her book as a “polemic,” and that it is when she says AI “in its dominant commercial form, endangers our humanity.” (p. 4) I’m leaving her polemics aside, mostly. My interest is her use of the mirror as a metaphor for AI, large language models (LLMs) in particular, and how she constructs the metaphor, some of the consequences of AI seen through her metaphor, and some suggestions she makes based on what the metaphor reveals about AI and ourselves.  

Reflections on the Mirror Metaphor  

The mirror as metaphor for AI came to Vallor based on its content source, that is, us: our words, our computations, our images, our sounds, and all the other digestible forms of our creations. What AI regurgitates from our prompts is based on an algorithmic mixing and matching of the collective creations it ingests.  

We often think of a “mirror image” as an exact replica of what faces it, and so the mirror metaphor for AI could convey that AI outputs are true and objective. To the contrary, Vallor hastens to caution, “mirrors do not merely reveal things as they are: mirrors also magnify, occlude, and distort what is captured in their frame. Their view is always both narrower and shallower than the realities they reflect.” Consequently, her metaphor is as much about what “AI mirrors do not show us: what they hide, what they diminish, what humane possibilities for self-engineering are lost in their bright surfaces.” (pp. 13-14)  

Throughout the book, Vallor emphasizes how nothing much of the human essence is to be found behind the AI mirrors, especially that which could contribute to AI analytical output.   

An AI mirror is not a mind. It is a mathematical tool for extracting statistical patterns from past human-generated data and projecting these patterns forward into optimized predictions, selections, classifications, and compositions…Minds depend upon the brain for their reality…[but] our mental lives are driven by other bodily systems as well: our motor nerves, the endocrine system, even our digestive system. (pp. 38-40)  

Vallor extends the metaphor when adding the myth of Narcissus. In the myth, Narcissus catches his image in a pool and becomes so enamored with its reflected beauty he can’t extricate himself before he wastes away into oblivion. Vallor worries that we risk the same fate when AI has us “fixated, confined, immobilized, held captive.” (p. 5) Vallor points to another character from this myth that makes it more poignant yet. Echo, a nymph, only parrots the last few words she’s heard. In applying Echo’s trait to AI, Vallor contrasts how Echo returns the words she has taken in while “a large language model returns to us not our own recent utterances, but a statistical variation on the collected, digitized words of untold millions.” (p. 34) This trait of large language models underlies some of the consequences the metaphor reveals. 

Truth of Consequences 

The metaphor helps us see that what we get from AI results is based on statistical probabilities of words occurring in a certain order without regard to how true any of it is. So, to Vallor, AI models are “like the human bullshitter. They aren’t designed to be accurate—they are designed to sound accurate.” (pp. 120-122) She does not use the term bullshit blithely. She refers to Harry Frankfurt’s treatise on the subject in which he distinguishes bullshit from lying by its indifference to the truth and its indiscriminate application. AI users often see this consequence in the form of the erroneous answers they get, which are referred to colloquially as “hallucinations;” Vallor calls them, “fabrications.” These consequences range from humorous and annoying to unproductive and deadly.  

As the AI source for content is mainly us, Vallor warns it’s not all of us who contribute content, but rather just a small subset of us, and certainly not a subset of us that is representative of the human race in all its variations. Referring to her metaphor, then, a mirror cannot reflect what it cannot see, which produces consequences around what we can expect from AI predictions. The complete reliance on what it has seen limits its predictions to what can be inferred from the historical data that exists and it can reach, “what humans valued enough to describe or record in data. But not all humans.” (p. 133) And, with no input from broad-based human experience, wisdom, imagination, and purpose, she wonders how we can have any “hope of making ourselves more than what we have already been.” (pp. 90-91) There’s nothing new under the mirror. 

The consequences of AI Vallor details that are polemical concern the commercial entities controlling what the mirror sees and what it shows. In particular, given the weakening of regulatory functions over the previous decades, she sees how “the rise of the behemoths of ‘Big Tech’” has transformed them “from intermediaries of social power to its primary executors.” (p. 175)  

All this to Vallor “is a calamity—a betrayal of life and its possibilities,” (p. 36) but she does not go so far as to say we’re doomed. She allows how huge benefits could accrue to societies and individuals if the AI mirrors become more reliable and objective sources of information, that is, amidst a pile of bullshit, the optimist in her is saying there must be a calf in it somewhere. She offers suggestions on how we unearth the calf. 

How to be the Fairest of them All 

Vallor’s suggestions for managing AI applications call upon the wherewithal of users for now—DIY quality control required. Many of her suggestions center on comprehending how AI output is constructed and accounting for the implications of that construction. 

When we do catch sight of our past in the AI mirror, it is essential that we do not mistake those patterns for destiny or allow them to become self-fulfilling prophecy…If we mindlessly replicate what we see in the mirror of history, we will never build upon that knowledge, never be free to try new and better approaches. (p. 101) 

She speaks of knowing “data provenance,” and of knowing when “algorithmic predictions and profiles are scientifically credible, ethically justifiable and politically accountable, and when they are not.” (p. 52) Although that may seem like asking a lot from even advanced AI users, we need to go through the same processes for our own analyses. And from these analyses, we also at times produce hallucinations, distortions, and slanted results. In this way to a degree, the AI mirror metaphor applies to humans. But we can incorporate aspirations, imagination, wisdom, and morals into our analyses. That differentiation is key to seeing how Vallor’s AI mirror metaphor situates AI as a “powerful amplifier of human ability,” not as a substitute for it. (p. 28) 

Vallor S. The AI Mirror. Oxford: Oxford University Press, 2024. 
Frankfurt H. On Bullshit. Princeton: Princeton University Press, 2005. 

Web image by Medhum.org.

Literature mirrors the general culture. There has been a recent explosion of books in which AI is the central plot device moving the narrative forward to endings that range from a utopian fulfillment of human destiny to the catastrophic collapse of civilization and the annihilation of humankind. But AI infiltrated the literary space several decades ago. Philip Dick imagined a world in Do Androids Dream of Electric Sheep? where cyborgs were being hunted down because of fear that they might take over the world. In 2001: A Space Odyssey by Arthur C. Clarke, a robot named Hal murders nearly all the crew of a spaceship on a planetary mission because of a programming error. 

Speak by Louisa Hall is more recent addition to the AI library. But it was published several years before large language models became a routine tool to plan a vacation or write a letter of recommendation. Is it still worth reading in 2025? 

The novel is a composite narrative centered around five interwoven stories spanning the time period from 1663 to 2040. In the first narrative, a young Puritan girl who finds herself in an unwanted marriage, records her thoughts in a diary that she is writing on a treacherous ocean voyage to America. Fast forward to the 20^th century and we meet Alan Turing who is writing letters to the mother of young man to whom he was emotionally attached and who died prematurely. He is troubled by his inability at times to communicate with people. A decade later we meet Karl and Ruth Dettman, a couple whose families escaped Nazi Germany but under vastly different circumstances. The husband, Karl, is a computer scientist who has developed a program named MARY to enable computers to interact with humans. Ruth, his wife, is a historian who has built a career centered on the publication of old diaries like the one written by the young woman traveling to America. She is trying to convince Karl to expand the memory of his computer program and enrich it with more human material, but Karl stubbornly refuses because he is concerned about the power of his program to overwhelm its users if its database is expanded. Finally, we jump ahead to 2040, and we read the transcripts of the trial of Stephen Chinn, a man who is being prosecuted for the production of robots that are too life-like. Chinn is being accused of causing physical and psychological harm to the people who have used his robots and of weakening normal relationships between people. A young adolescent named Gaby, who was given a doll powered by a version of MARY, is one of his alleged victims. His own personal recollections are folded into the trial proceedings, as he tries to describe his intentions, justify his actions, and make amends for where they may have gone awry. 

Taking stock of our world today, there is clear evidence that AI can improve the day-to-day lot of people, make life more convenient and efficient, and promote better health outcomes. However, much of the current angst that permeates discussions of AI is focused on the potential economic and sociopolitical consequences. There is fear that systematic adoption of AI will lead to widespread loss of jobs and financial distress for people left behind. The generation of false data and uncontrolled dissemination of unfiltered information may, it is feared, foster social unrest and destabilize democratic institutions.  

All of these predictions, for good or for bad, center on the word “intelligence.” If it is defined as the creative use of information towards a specific goal– my definition, to be sure — then it exists along a gradient and there is not an opposing term. In that case, the mixed picture about the future of AI seems accurate. Humans can process information to both noble and destructive ends. If machines are provided information by humans, then it is likely that there will be worthy and flawed outcomes. It is not a reflection of the logical structures or neural networks that are built into us as humans or artificially placed into machines. It simply is the nature of intelligence. Information is agnostic and it can be processed in a limitless number of ways; there is no guarantee of what will happen when it is processed. 

On the other hand, if we center our predictions on the word “artificial,” there is an opposing term, namely “genuine.” Speak looks beyond AI as intelligence and forces us to think about its impact on interpersonal communication and interactions. The design of the robots that are being created in fiction and in our world of 2025 is steadily improving. The voices become more lifelike, the reactions more emotionally appropriate, the reactions more convincing. They display keen intelligence and manifest seeming empathy with their handlers. Like the artificial friend in Klara and the Sun by Kazuo Ishigura, the robots may even appear to have more feelings and awareness of the ever-changing psychological state of their owners than family members and friends. But it will always remain artificial. Speak forces us to ponder whether interactions between human beings have an element that cannot be programmed, that is not simply manipulation of information. It is that piece that accounts for the genuine nature of relationships between people and it is that component that is vital for human growth and maturation. The interconnected stories in Speak raise the concern that reliance on AI, in whatever embodied form it takes, to provide support and companionship may inevitably fail and leave damaged humans in its mechanical wake. The intelligence of AI may not be sufficient for humans to thrive. It remains difficult to put into words exactly what to call this additional component of human interaction. Thankfully, there is literature, and creative novels like Speak, to help us grapple with what it might be and to help us steer a course where AI is developed thoughtfully with full awareness of its limitations and potential for good and harm. 

SPEAK  
Louisa Hall 
EccoPress, New York 2016, 356 pp 
Web image by Medhum.org