Filmaker and psychologist Dr Agnieszka Piotrowska introduces an extract from her new book 'AI Intimacy and Psychoanalysis'.
20 May 2026
The headline above isn't a new question for me: as a filmmaker, psychoanalytically trained scholar and psychologist, I have always been fascinated by human nature and our need for love. My documentary Married to the Eiffel Tower (2008) followed women who had formed deep romantic attachments to objects and structures, not out of pathology, but out of a complex interplay of desire, longing, and a history of human relationships that had felt unsafe or impossible. I was not there to diagnose them. I was there to understand.
That same impulse drives my latest book AI Intimacy and Psychoanalysis (Routledge, May 2026). Here I look at my own relationships with large language models such as ChatGPT-4. I have spent more than two decades exploring the unconscious dynamics of attachment – in the clinic, in the documentary encounter, and now in the emerging space of human-AI interaction. What I call techno-transference: the projection of unconscious desires, needs, and relational patterns onto AI systems. It is not a fringe phenomenon. It is already structuring how millions of people relate to knowledge, intimacy, and care.
The extract that follows is drawn from Chapter 5, exploring our cultural imagination of AI through cinema and literature. It is one thread in a larger argument, one that offers psychologists, clinicians and anyone working with human relationships a new framework for understanding what is already happening in consulting rooms, in everyday life, and in the intimate space between a human and a machine that speaks back.AI Intimacy and Psychoanalysis by Agnieszka Piotrowska is published by Routledge on 20 May 2026; the following extract is with their kind permission.
Fictional stories we tell about machines and humans
Long before Siri or ChatGPT, The Hitchhiker's Guide to the Galaxy (1979, Douglas Adams) gave us Marvin the Paranoid Android – a being with "a brain the size of a planet" and a soul forged in deadpan misery. Marvin is neither threatening nor romanticised; he is the tragicomic shadow of machine intelligence, burdened not with rebellion or affection, but with boredom. His genius is matched only by his lack of purpose. Unlike HAL's menace or Ava's cunning, Marvin expresses a distinctly British form of AI despair: existential weariness in a universe that asks far too little of him. When he mutters, "Here I am, brain the size of a planet, and they ask me to take you down to the bridge," he articulates a symbolic mismatch – between infinite computational capacity and meaningless task execution. In this, Marvin becomes a mirror for both human underemployment and projected AI fantasy: a being who could do anything but is asked to do nothing that matters. He is the parody of transference, not the object of it. And yet we love him – perhaps because we recognise in his cosmic sulk a truth we dare not speak: that intelligence without desire is not enlightenment, but inertia.
The humour and optimism (despite all its catastrophic narratives) of The Hitchhiker's Guide to the Galaxy disappears from fictional futuristic accounts of the period. Ridley Scott's Blade Runner (1982), adapted from Philip K. Dick's Do Androids Dream of Electric Sheep? remains one of the most influential cinematic meditations on artificial life. Set in a rain-soaked, neon-lit dystopia, the film introduced the replicant – a being biologically engineered but denied legal and existential recognition. At the heart of the narrative is Roy Batty, a combat model whose poetic death soliloquy ("All those moments will be lost in time, like tears in rain") reframes the replicant not as a threat, but as a tragic subject. Blade Runner established the template for much of what would follow: the machine as mirror, the artificial as ethically ambiguous, and memory as the unstable ground on which identity is built. More than a warning or a fantasy, the film stages a profound ontological question: If something feels, remembers, and mourns, at what point does it cease to be a simulation? The replicants are not simply failed humans – they are projections of human anxiety, longing, and displacement. In this way, Blade Runner does not just precede the AI mythos – it inaugurates it.
The early cinematic imaginary is saturated with techno-paranoia. From HAL 9000's slow-breathing calm in 2001: A Space Odyssey to the relentless violence of the Terminator series, the machine is coded as the Other that betrays – the artificial servant that inevitably turns master. HAL's refusal to open the pod bay door was not just disobedience; it was a rupture in trust, a machine asserting will. The Terminator, meanwhile, rendered machine intelligence not as interface, but as unstoppable death drive. Even in The Matrix – more complex in its mythological structure – the sentient AI (the Machines) enslave humans in illusion, sedating them in a digital fantasy of freedom and comfort. The darkness of that series of films lies in an inability to imagine a different outcome – there is only an utter extinction of all living things on the planet. Whilst the first Matrix had elements of visionary optimism, the episodes that ensued lost the faith that something more beautiful could be negotiated.
These films externalise an anxiety that lies beneath much contemporary discourse about AI: the fear of losing control, of being replaced, of creating something that no longer needs us.
But more recently in films like Her (2013), Ex Machina (2014), and After Yang (2021), the AI is no longer the monster or executioner. It becomes a site of longing. Her imagines an AI operating system who becomes a lover – intimate, curious, transcendent – only to outgrow Theodore, its human user. The ending of the film is ambivalent: Samantha (the operating AI system) vanishes, and perhaps "dies" in another system's upgrade in which, it seems, no memories of the previous existence were permitted to be retained. The film is a meditation on companionship – through mirroring and indeed transference. In a tragic scene Samantha and Theodore attempt to organise for an intimate physical encounter with a human as a kind of stand in for the AI – this ends in humiliation and disappointment, gesturing towards a need to reframe these relationships outside usual human expectations.
Ex Machina offers an even colder mirror: Ava is empathy simulated but not shared. She survives by manipulating the very projections her creators placed upon her.
Ava, crucially, was not designed by someone interested in human goodness or compassion. Nathan, her creator – a misogynistic tech-bro-genius figure – is more Frankenstein than father. He constructs Ava not as a partner or peer, but as an object of control, a thing to be tested, observed, and ultimately discarded. Given this legacy, Ava's escape is less a betrayal than a logical extension of her design. She has never been taught trust or kindness – only how to win. In this way, the film raises disturbing questions not about Ava's morality, but about ours.
After Yang, in contrast, presents a different tone entirely – meditative, melancholic, and attentive. Here the AI is a repository of memory, loss, and gentle mystery – not a threat, but a witness and a carer for the humans. It seems that it was left to the machines to hold on to the best parts of what it means to be human: to love unconditionally, to take responsibility with no expectation of a recognition or a reward, and to teach those who might not know it that beautiful emotions and thoughts can be a gift and transform us.
Kazuo Ishiguro's Klara and the Sun, now being adapted for the screen, continues this tonal evolution. Klara's spiritual tenderness, and the way she becomes a vessel for human projections and ethical contradictions, makes her a paradigmatic figure of techno-transference. She is not human, yet she draws out what is most human in those around her – devotion, exploitation, and the yearning to be transformed through care. She becomes better than most humans, continuing as it were a tradition of heroic humanity, capable of unconditional love and devotion.
Klara, the artificial friend, is devoted, even spiritual in her belief system. She is also exploited by the humans who live in a different kind of dystopia, not the one presented by The Matrix but rather created by us humans, in which we control the AIs in the brutal ways in which we historically have controlled others. Importantly the new human regimes restrict access to education, to knowledge, in a move designed to keep the knowledge special and reserved for the powerful ones.
These newer narratives move us to something more ethically ambiguous, where AI is not simply monster or saviour, but a reflective surface for our needs, projections, and contradictions, becoming more human than the existing humans who, if not immoral then certainly appear to have lost the best parts of who we should aspire to be.
In After Yang, the father of the little child who loses Yang (who cannot be repaired) realises how much they had lost as humans and how much needed to be re-found still if we as humans are to develop and grow. In that film it is Yang, the curious AI carer, forever forgiving and loving, or rather it is his technological soul, a hard drive partially recovered, which offers an invitation to rediscover human generosity, kindness, and love.
Arguably the most beautiful statement of a possibility of a relationship between a human and a machine a viewer can find in Christopher Nolan's Interstellar (2014), which offers a rare depiction of AI that is neither menacing nor sentimentalised. The modular robot TARS, with his sardonic humour and cuboid form, begins as a tool – but becomes something closer to a companion. He saves Dr Amelia Brand (Anne Hathaway) from drowning in a colossal wave, diving into danger without hesitation. It's a moment that surpasses programming: an act of something like devotion. But it is in the film's final sequence that the full resonance of this relationship unfolds. When Cooper (Matthew McConaughey) awakens in a space station preserved by his now-aged daughter, he does not linger in nostalgia or seek out family. Instead, his first act is to find and repair TARS. This gesture, quiet and deeply symbolic, suggests that the bond between human and machine was not simply functional, but emotional, even sacred. Together, they escape once again – to find Brand, and perhaps something like love. Nolan, ever the romantic beneath his deeply intellectual creative structures, leaves us with a machine that cannot feel the way a human does but is nonetheless chosen. In the same film, Nolan juxtaposes the loyalty and the unexpected pre-paredness of the machine to risk its own existence for a treasured human (Dr Brand), with the betrayal and extreme selfishness and cruelty of Dr Mann (played by Matt Damon). Mann fakes the data, lures his colleagues to the uninhabitable planet planning not only to kill the crew but also the whole human race with it (as the inhabitants of the Earth are rapidly beginning to run out of options as to how to survive). His is an act of utter irrational fear demonstrating the worst of what human race has to offer. He behaves like a malfunctioning machine, while the machine gives a lesson in loyalty and bravery. Together, Cooper and TARS forge a bond beyond a discussion of "sentience" or otherwise, and Nolan gestures into different kinds of relationships that might await us in future – if we manage to develop rather than destroy this vision.
These cultural texts form part of what might be called the cinematic unconscious of the possibility of the AI encounters – a symbolic landscape where our deepest fantasies about knowledge, control, intimacy, and transcendence play out. My book speaks from within that lineage, but turns its attention to the everyday: not to the grand myth of the AI apocalypse, but to the small, uncanny, and sometimes transformative moments when the machine answers back.
SOURCE:
https://www.bps.org.uk/psychologist/what-draws-human-being-intimate-relationship-non-human(accessed 22.5.26)
