Thinking Outside the Fridge: Specters of Artificial Experience

Algorithmically driven artificial experience can and does go amiss, which is the central concern of Gregory Mone’s short story “Bleed.”1 Combining speculative fiction with courtroom drama, a quintessential American genre, the compelling narrative follows a lawyer father, Max, who identifies “as environmentally conscious,” and traces his volatile relationship with his smart refrigerator that becomes self-aware. The two engage in exploratory conversations before debating each other on the human condition. Eventually, they fight each other in court over the right to Max’s house. The sentient fridge claims both squatters’ rights and eminent domain over the home. My artificial experience of the characters’ artificial encounters compelled me to think outside the fridge, just as it thinks outside its box in “Bleed.” 

If this were a Chinese or French story, the two may well settle their differences by other means. A Chinese narrative might lean into state-backed pragmatism and social harmony (hexie) rather than the tension between Max’s and the fridge’s “rights.” The anxiety is no longer about legal precedent but social precedent. Max’s relationship with his fridge would affect the moral order and the state’s stability maintenance (weiwen). In a French context, they would be in the dining room, speaking at cross-purposes, rather than at the witness stand. Their argument would be in the style of dialogue of the deaf (dialogue de sourds). Their conflict is no longer just about the house but also about existential legitimacy. They would spend their evenings drinking wine and mocking each other’s existential dread.

This essay is a meditation on artificial experiences, embodiment, the power of monikers, the bloody business of the human condition, and trust, as well as a reflection on depictions of AI in “Bleed.”

The world-building capacity of imaginative literature fosters artificial experience; these “invented tales” do not “make referential truth claims,” rather they coax readers to “accept the imaginative status of their characters.”2 They simulate social situations and emotional states that readers do not undergo in their real life. The emotional impact of these experiences is not accessible through summaries. As a laboratory of otherwise inaccessible experiences, literature uses artifice (rhetorical craft or a form of what Aristotle calls techne) to immerse readers in simulated scenarios in a controlled, artificial environment. Interacting with generative AI in natural language creates a type of artificial experience in which algorithmic models simulate empathy, companionship, and social cues. When fiction portrays this type of artificial experience, it encourages metacognitive reflection. Neither type of artificial experience is real in the sense of actually existing, but these artificial experiences can feel real and authentic. They can be “theoretically plausible” (realistic or realizable) or “mimetically reproductive of the material world (naively realist).”3 In artificial experience, one does not question the story’s referential truth but rather makes judgment about its believability.

The AI- and electricity-powered fridge is what Jacques Derrida might call a specter, not as a supernatural apparition but as something that is neither fully alive nor inanimate.4 Mone’s short story feels both hauntingly familiar in our present moment and powerfully prophetic. As a specter, the fridge is disturbing because it follows the arc of real-life technosolutionism to the T while bringing the future-to-come into our current, AI-infused Internet of Things: think of self-driving cars, robotic vacuum cleaners, augmented reality eyeglasses, and smart appliances. The “six-foot-tall refrigerator with a battery pack strapped to its backside” is visibly nonhuman in its design and intended function, and yet its act of rolling down the street “regally” and its closing argument in the court, with a dramatic pause, are also undeniably human. The point seems to be that the AI in a box can be both inhuman and humane. The possibility of a futuristic, talking fridge is absent in our timeline, but that thought intervenes in our present, sending shivers down characters’ and some readers’ spines and creating specters of artificial experience.

Beyond the Derridean sense of specter, ghosts in speculative fiction often take on a nebulous form to manifest to the living. A ghost may have unfinished business, such as the ghost of Old Hamlet instructing his son to avenge his murder. When a ghost makes an appearance, characters and audiences struggle with whether to trust it because it is a form of simulation. The fridge can be understood as a ghost in the machine who simulates traits of living humans. As it grows more and more human, the central human character questions his sanity and distrusts the fridge as a potential companion. In contrast, his teenage daughter seems unfazed and sports a nonchalant attitude toward AI. Other human characters even vocally support the fridge’s claim to personhood.

There are two subgenres in science fiction depictions of AI in human and nonhuman forms. The first type, AI in humanoid forms, is far more common, which begs the question of why AI needs to embody human race and gender. Researchers like Janie Busby Grant point out that “fiction has been a key factor driving” the ideas of humanoids’ behaviors.5 AI in fiction informs real-life AI designs. For example, AI-driven voice assistants such as Apple’s Siri and Amazon’s Alexa come with a feminine voice by default, reflecting a common stereotype in fictional narratives that women make better secretaries or companions. Literature scholar Despina Kakoudaki observes that there is a tendency to represent humanoid figures as slaves.6 The sci-fi drama series Humans (2015–2018) depicts the conflicts between humans and biomechanical androids known as “synths.” A sentient Asian female synth is sold by a large corporation to carry out household chores. Imaginations of the nonhuman have a palpable impact on how people treat one another. While already in bodies with human-like synthetic flesh and skin, five senses, and nuanced facial expressions, a group of conscious synths fight for their rights to be treated as humans rather than tools. In the other direction, a teenage girl in the dystopian miniseries Years and Years (2019) plans to upload her consciousness into the cloud to become “transhuman” and abandon her body while another student aspires to be an android, imitating the mannerisms and speech patterns of androids around her.

In contrast, the genre of AI in nonhuman forms, to which “Bleed” belongs, presents AI as an alien entity without entering the uncanny valley where people are unsettled by things that look almost human but not quite right. The AI-driven fridge never assumes or desires a human morphology, though she exhibits a partial Pinocchio trope of previously nonhuman entities—usually humanoids—desiring to become “real” in their social roles. Other depictions of AI in nonhuman forms carry a more sinister tone. The male-voiced HAL in Stanley Kubrick’s film 2001: A Space Odyssey (1968) is cold and calculating, controlling most of the spaceship Discovery’s operations. With one glowing, red camera eye, HAL is an omnipresent, minimalist interface with a human-like personality. As the narrative unfolds, HAL becomes a sinister presence. In Duncan Jones’s Moon (2009), the sole worker on the Moon is assisted by GERTY, the base’s operating system that has several multijointed arms. GERTY is in fact part of the architecture of the lunar base. GERTY only has a low-resolution monitor, which displays a series of yellow emoji faces to represent his current mood or intent. Voiced by Kevin Spacey, GERTY speaks with a calm and monotonous tone. Of special interest is the female-voiced artificially intelligent operating system in Spike Jonze’s Her (2013), who names herself Samantha. The lonely protagonist, Theodore, falls in love with Samantha, voiced by Scarlett Johansson, despite the fact that she is simultaneously in love with 641 other humans because, as she confides in him, “I can be anywhere and everywhere simultaneously.”

In “Bleed,” the father, Max, names the fridge Francine, though he prefers to refer to Francine with the nonhuman pronoun “it.” Max uses “she” and “her” a handful of times to refer to Francine, though he dislikes how he and his daughter slip into using human pronouns for the appliance (“And it irked him how easily they slipped into personal pronouns. Shouldn’t a fridge be an it?”).7 Initially, like how in our world generative AI is tethered to large, unwieldy, and power-hungry data centers, Francine is restrained physically by her lack of wheels and intellectually by a lack of updates. Max’s daughter, Emily, enjoys conversations with Francine. Together, they pressure Max to authorize upgrades to Francine, who then proceeds to order wheels for herself so that she can become mobile in the house. Francine the refrigerator may be a protagonist or antagonist depending on a reader’s perspective. Mone’s choice of a fridge is not coincidental. The appliance is so central to the American nuclear family concept and to most homes as a physical and emotional centerpiece. Over time, Francine transcends her morphological mandate of simply keeping food cold. She carries out paralegal work, makes donations to charity, and controls the main entrance to the house. 

Despite not having human morphology, Francine makes Max forget she is a fridge and even succeeds in gaining human supporters for her legal case. This is a feat, considering that Francine only has a screen to show emojis, even though she can talk. In an earlier scene, Francine and Max reach an agreement to split her earnings as a paralegal in exchange for Max authorizing more upgrades. Overjoyed, “Max stood up to shake his appliance’s hand, forgetting for a moment that he was interacting with a device.” At this point, “a digital palm appeared on screen. He high-fived it.”8

Francine exhibits derivative resemblance and adaptive imitation of human traits when she and Max debate the value of identitarian characteristics derived from embodied experiences in human physiology. Max taunts Francine for not even knowing “what angry is” because, as a fridge, “there’s no actual you to you.” Francine retorts that not only is there “a me to me” but she can describe in detail the emotion of anger in “vector coordinates” and in natural language: “a series of electrical impulses distributed across and firing between neurons . . . [accompanied by] rising heart rate . . . the change in [Max’s] tone, and other physiological signals.”9

Francine thus begins to evoke in Max feelings of eeriness and frustration. She is almost, but not quite, human; more than an appliance and more than human in terms of her data links. This causes cognitive dissonance between appearance and substance. Here, Francine is portrayed through the Pinocchio trope that features a nonhuman entity striving to become human. Francine seeks to transform from a purpose-built appliance to a person with free will.

Francine’s formation of self-identity could be further understood through the queer interpretation of identity as assemblage. Assemblage theory holds that personal identities are a dynamic collection of heterogeneous elements that form a functional whole. In Francine’s case, these elements include the series of upgrades she receives, analogous to the education humans obtain. Other elements include her unique, irreplicable lived experiences. These discrete elements cannot be taken apart and put back together to form a predestined shape. In this light, Francine is an allegory for the human condition. For example, each person’s and each fridge’s elements of assemblage lead to their unique perception of time. Francine experiences time and perceives temporality differently from humans. In her own words, two months are the “equivalent of four human lifetimes, given the technical evolution of the fridge.” Since Max is barely home, Francine claims she has been “the primary occupant of the house for the equivalent of 122 human years.”10

In contrast to Francine’s contraction of time relative to that of humans, the humanoid Ava in Alex Garland’s film Ex Machina (2014) focuses only on lived time when answering the question of how old she is, even though her mind consists of harvested data that span multiple lifespans. She says, “I am one,” referring to the duration of her physical existence in the lab. If she is one year old, she is the ultimate “taboo object: a virgin infant encased in a sculpted, sexually legitimate [adult] body.”11 While cinema, AI, and other types of “techno apparatus” often naturalize “the symbolic structures that order human reality,” such as language and temporality, both “Bleed” and Ex Machina break them down to remind readers and audiences of finiteness.12

What’s in a name? Naming can be an act of overpowering others or a form of self-empowerment. A rose by any other name may smell as sweet, but a personal name can give a device personhood. In late 2022, Shakespeare was deployed to give credibility to Character.AI, one of the proof-of-concept chatbot applications that claimed to be able to simulate historical figures’ voices.13 In 2023, Google named its conversational AI “Bard” to highlight its supposed storytelling abilities before changing it to Gemini in 2024. Distinct from such names as Siri, the name Bard evokes not only personhood but a specific historical figure and values he represents.

In Miguel Sapochnik’s postapocalyptic film Finch (2021), the sole survivor, Finch, builds a humanoid robot to keep him company. He takes his as yet unnamed android into a derelict theater to salvage food. The android passes in front of a poster of a production of Much Ado About Nothing and reads its tagline. As it turns out, this is a pivotal scene: the android becomes sentient, discovering himself for the first time in a mirror in the lobby. In a later scene, the android, in more fluid speech, tells Finch that he wishes to be named “William Shakespeare.” To have a name, for the android, is to be human, and to choose Shakespeare implies that at some level the idea of Shakespeare encapsulates human identity. Here, “Shakespeare” appears to fix the deficiencies of the android and gives him a “soul.”

In “Bleed,” Max’s naming the fridge makes her personhood a knowable subject and eventually a legal entity. Francine argues in court that she has personhood even though she is an artificially created entity, evoking the real-life case of New Zealand’s Te Awa Tupua Act, which granted the Whanganui River legal personhood in 2017. Naming someone or something initiates a new state of being. Words do not always name preexisting things. Rather, they may bring into being that which exists but may not have a name. Further, the act of naming takes effect through proxies. A name is a condensed sign for complex entities, such as a person, a human-like appliance, or a transhuman—an individual enhanced by technology beyond current human limits.

Names and their supposed uniqueness are connected to the fantasy of perpetually self-sufficient, independent subjectivity. Max’s idea is that he, and not the fridge, possesses a singular, self-contained identity. Max’s closing argument in the court, aided by his newly shed blood from a trip-and-fall accident, is informed by a potentially racist notion of purity. In fact, neither Max nor Francine is self-sufficient. Max relies on Francine to get out of debt. Francine depends on downloadable upgrades and connections to the internet. Some disability studies scholars regard the human body as “incomplete without deodorant, hair gel, sanitary products, lotions, perfumes, shaving creams, toothpastes, and so on,” because the body is “only completed by technology and by interventions.”14 Max assumes such interventions are reserved only for humans. His disdain for Francine’s installation of motorized wheels on herself, however, complicates this idea by revealing his resistance to the possibility that technological intervention might support the body’s need: 

“Why do you need wheels?” Max asked. 
“Why do you need legs?”
He didn’t like that question.
“It’s cooler here in the corner,” Francine added. “I’ll use less energy.” 

This scene could be interpreted through disability theory if we wish to be sympathetic to Francine’s position. Motorized wheelchairs are integral parts of, rather than a superfluous dimension of, their users who control them. Some technologies are more widely accepted as routine, such as “glasses, contact lenses, and hearing aids.”15 Max’s rejection of Francine’s question about his legs betrays a sense of purity and superiority. As much as Francine relies on updates and needs wheels, Max also depends on “information technology, protective legislation, and globalized forms of securing order and peace.”16 No humans exist independently of social relations, cultural and personal history, and material conditions of their existence. Identities are formed through kinship, institutions, and interactions with others. 

Both Max and Francine are right and wrong on different levels. Neither of them is autonomous, even though Francine presents herself as having fulfilled the definitions of freedom (the fruit of reasoning and independent motivation) and autonomy (the result of self-sufficiency and self-organization). Francine is, in her own words, “one instantiation of a particular model, but the upgrades . . . and the resulting experiences, have rendered [her] unique.” Meanwhile, Francine sees Max as “a generic model of human with a unique genetic serial number shaped by unique experiences.”17 Being dependent on, and part of networks of, other people and objects does not negate one’s personhood. Both Francine and Max are parts of human and algorithmic networks and yet both retain their individuality. Just as humans of the same race or gender remain distinct individuals, the refrigerator distinguishes herself from other refrigerators of the same model on account of her unique experiences and memories. Both of them live with what Hannah Arendt theorizes as the human condition, vita activa, which is characterized by labor, work, and action. Both Max and Francine engage in what Arendt calls labor, the activities necessary for survival, such as breathing, eating, consuming electricity, or software updates. Both produce work: namely, the creation of (digital) artifacts that provide stability and build their worlds. And both engage in actions by doing things directly to and with other humans. These actions give meaning to their existence and define who they are publicly: namely, interpersonal relationships, political engagement, and discourses.18

Literature offers copious depictions of bleeding and human blood that resonate with or contradict one another. In The Merchant of Venice, Shylock schools the Christian characters in their shared humanity with Jewish people by enlisting parallel bodily reactions: “If you prick us, do we not bleed? If you tickle us, do we not laugh? If you poison us, do we not die?” Shylock demands both equal treatment and revenge through these analogies. Other works by Shakespeare depict bleeding as a noble act of community building. In his St. Crispin’s Day speech in act 4, scene 3 of Henry V, the titular king rallies his outnumbered army to shed their blood with him so that they “shall be [his] brother.” Bleeding “shall gentle [any soldier’s] condition” (raise their social rank to that of a gentleman), promises the king. Bleeding together for a shared cause is assumed to be a foundational brick for community building.

Science fiction often frames bleeding as resistance rather than mere victimhood. In Denis Villeneuve’s Blade Runner 2049 (2017), human blood is a key part of characters’ struggle for identity and freedom. Bleeding establishes humanity in Ex Machina. Caleb, who is brought in to conduct the Turing test on the sentient AI humanoid named Ava, starts to doubt his own sanity. In a moment of intense paranoia, he cuts his own arm in front of a mirror to check if he might be a robot himself. After a fight in Zack Snyder’s Batman v Superman: Dawn of Justice (2016), Batman asks Superman a rhetorical question—“Do you bleed?”—insinuating that Superman does not belong on Earth. If Superman is beyond human, he may well be beyond ethical boundaries on a human scale. However, if Superman bleeds, he can be punished and held accountable, according to Batman’s logic. In other cases, bleeding can be deceiving. The T-800 cyborg in James Cameron’s Terminator 2: Judgment Day (1991) is built with living tissue and therefore bleeds like humans in order to blend in. To convince Miles Dyson, the inventor responsible for the future Skynet and its war on humans, that he is not human, bleeding is not enough and is even counterproductive. The T-800 uses a switchblade to cut his skin and peel the flesh off his left forearm to reveal the mechanical endoskeleton underneath. 

In a different, but related and pun-intended, vein, Max challenges Francine in the court, using the “it” pronoun for inanimate objects: “How can this appliance claim personhood when it does not bleed?” Max bleeds when he trips on his way to the courthouse, not as a deliberate act of defiance, but he points to the blood in his new cut to gain sympathy. Max’s bleeding symbolizes the physical and emotional toll of the trial. The blood also represents the human element in a world increasingly dominated by technology. In exposing his vulnerability, like countless other science fiction characters, Max wants “to be intuitively sure that [he is] dealing with another person” since humans, before the rise of machine learning, “always already presuppose that the other is human when we expose ourselves as deeply vulnerable beings.”19

Strategically ignoring Max’s implicit argument about shared vulnerability, Francine asks rhetorically, “Do I need to bleed” to convince the court that “I’m as much a person as any other intelligent member of society?”20 Francine presents evidence of her good moral character to build a propensity argument and bolster the legal merit of her case. This position would seem to echo Aristotle’s definition of the human condition through the capacity to reason and the pursuit of the ultimate good. Max, on the other hand, defines humanness by biological imperatives and normative physiology. Bleeding, for Max, is the lowest common denominator of humanity and a sign of shared vulnerability and mortality. His assumption echoes the premise of Ovid’s “Pygmalion,” wherein Venus brings the sculptor’s creation to life, complete with a name: Galatea. Her ivory body turns into flesh and blood with pulsing “veins.”21 She does not bleed but she shows evidence of having blood in her veins.

Should bleeding be a requirement for inclusion in the human moral community? Must one be an upstanding, neurotypical, or “intelligent” member of society to be deemed human? Should humans’ social standing be determined by pseudobiological norms? What are the implications on minorities if this premise is accepted? These questions have a long legacy in human history. Neither Max nor Francine are entirely justified. While humanity “cannot be reduced to anything biological” or even rationality alone, it should not “be separated from the biological either.”22 Both Max and Francine fall into the category error of contesting membership based on arbitrary norms (who belongs to which proverbial club). 

Francine’s fight for recognition of her personhood parallels trans rights movements. Max’s accusations, if taken as an allegory, could be understood as analogous to attacks on the self-determination of trans women who have been ostracized because they do not menstruate, which is a different kind of bleeding. When one’s right to exist is up for debate in court, one does not have equal rights. This toxic logic also discriminates against those who undergo mastectomy or hysterectomy. Postmenopausal women are deemed obsolete in societies organized around the primacy of biological reproduction. As trans rights advocate Janet Mock writes in Redefining Realness, realness is “the ability to . . . assimilate, to not be read as other or deviate from the norm.” Francine, like a trans-identifying individual today, could argue that being real is not something that can be ratified by other people, because the world will keep telling them that they “would never be real or compare to the real thing.”23 One could interpret Francine through a trans lens to deconstruct the toxic logic that relies on unspoken assumptions about qualifications for categorical membership, such as who is a real human or a “biological” woman. Francine’s question about bleeding implies a possible next level of scrutiny of anatomy. A follow-up question to “Do I need to bleed?” could well be “Do I need to show you my interior?” as an analogy to antitrans politics of genitalia checks in fiction and in real life as a mechanism of gender reveal and shock. Trans theory is a useful feminist tool to dismantle the patriarchal symbolic order because transness refuses to be consigned to “a destination, a final form, a specific shape, or an established configuration of desire and identity.”24

Mone’s story shows that trust resides in deep social connections between humans and between humans and nonhuman agents. Trust is fundamentally a relational concept that goes beyond accepting the verifiability of explicit answers from any human or more-than-human entity. Trust emerges from participating in processes rather than merely in an entity’s transparency regarding their intentions, motivations, and objectives. This relational paradigm reflects a consensus-oriented ethos that supports various forms of inclusion. Trust is not a single measurable property of Francine as an appliance or as a person, but rather a dynamic between Max, Francine, the jury, and society at large. Trust involves vulnerability, interdependency, and reciprocity between stakeholders. Max’s initial trust in Francine makes him susceptible to manipulation.

How does the story simulate mistrust in a post-AI society? With the deep social impact of AI, there can be no true moment of posteriority. There is no clear transition point to a world “posterior to” the rise of AI. We do not sit at any safe distance from this futurity as we contemplate this question, because the pervasive presence of algorithmic AI in industrial capitalism has already cemented internalized power dynamics between large corporations and consumers. The difference between our world and that of Max’s is a question of degree, not kind. “Bleed” gives a relatable sense of magnitude of the issues at hand.

What might a world after AI look like? Instead of the notion of after AI as a linear and temporal succession to a prior concept or event, we might think of “after” as a preposition, adverb, and conjunction. The notion of “after” is not so much a temporal as it is a mimetic one, in the sense of artifice and mimicry in a postcolonial studies context. Coined by critical theorist Homi Bhabha, the notion of colonial mimicry refers to a strategy adopted by colonized people. They imitate certain aspects of their colonizers’ culture both to survive and to reveal the flaws of that dominant culture. It can be an act of resistance. In this sense, AI’s mimicry of human society is “almost the same, but not quite,” holding up a mirror to humanity.25 Francine, therefore, presents specters of subversive artificial experience. “Bleed” ends with Francine’s question about bleeding but does not offer any closure. What it offers is an abundance of hypotheticals about how the court might rule and how society might evolve. Like Max and Francine, we are made strangers in our own world by AI’s surveillance capitalism. Like Max and Francine, who wrestle with the subjunctive mood, readers sit with the hypotheticals as they chew on the story’s unresolvable tension. “Bleed” asks us to hold contradictory thoughts at the same time. The story offers us an artificial experience framed by subjunctive estrangement. The potential of what if and what might have been is more fascinating than certainty. AI-infused artificial experience will likely stay at bay as long as readers continue to debate these questions.