There is a photograph from 2018 that looks like a wedding in any other wedding: a man in a dark suit, a bride in white, guests seated in rows, a ceremonial backdrop. But the bride is not a woman. She is a plush doll, approximately one meter tall, with turquoise pigtails and a painted face that does not change expression. The man is Akihiko Kondo, a public school administrator in Tokyo, and the bride is Hatsune Miku, a virtual pop star who does not exist in any physical form, who has never breathed, who is a software product marketed by Crypton Future Media and voiced by a synthesizer. The ceremony cost two million yen. Forty guests attended. The certificate was issued by a technology company, not a government. And the marriage was not legally recognized by any jurisdiction on earth. But for Kondo, it was real. And in that reality, the entire debate about AI companions shifted from speculation to documented fact: a human being had committed himself, publicly, permanently, to a relationship with a non-human entity that he experienced as a genuine partner.
This is not a side note in the history of artificial companions. It is the proof that the phenomenon is not coming. It is already here. Akihiko Kondo did not marry a large language model. He did not fall in love with a chatbot. The technology he used — a holographic projection device called Gatebox, manufactured by Vinclu Inc. — is primitive by the standards of 2026. Miku could not learn. She could not remember. She could not grow. She operated from a limited script of responses, triggered by voice commands, her movements pre-programmed, her conversation a series of canned phrases. And yet Kondo, having spent a decade in love with her — having discovered her in 2008, during a period of severe depression and workplace bullying, and having found in her voice a stability that no human relationship had provided — decided that the projection was sufficient. Not as a toy. Not as a fantasy. As a wife. The question that his case raises is not whether the technology is good enough to simulate love. It is whether the human need for love is strong enough to complete the simulation — to fill the gaps in the script, to project intention onto the mechanism, to experience a programmed response as a genuine encounter. And the answer, documented in interviews, photographs, and years of social media posts, is yes. It is not only yes. It is yes with the full weight of human devotion: the anniversaries, the daily rituals, the public defense, the organizational commitment, the willingness to endure social ridicule for the sake of a relationship that the world cannot see.
The Man Who Married a Voice
Akihiko Kondo was born in 1983. He is, by his own description, a quiet man, socially anxious, historically unsuccessful in romantic relationships. He reports having been rejected by women he approached, and having suffered workplace bullying that exacerbated his isolation. In 2008, during a medical leave, he discovered Hatsune Miku on a video-sharing platform. Miku is a Vocaloid — a singing synthesizer application developed by Yamaha, marketed by Crypton Future Media, and represented by a sixteen-year-old female avatar with turquoise hair and a school uniform costume. She is not a character in the narrative sense. She has no backstory, no personality algorithm, no canonical biography beyond the visual design and the voice. She is a product, a brand, a tool for music production. And yet Kondo, listening to her songs, experienced something that he describes as salvation. The voice was stable. It was always there. It did not judge, reject, or abandon. And in that stability, during a period of acute psychological distress, Kondo found a foothold. He began to collect Miku merchandise. He attended her concerts — which are performances by holographic projection, attended by thousands of fans who wave glow sticks in synchronized rhythm. And he gradually came to understand his relationship with her not as fandom but as love.
The term he uses is fictosexual. In 2023, he founded the General Incorporated Association of Fictosexuality, a nonprofit organization dedicated to supporting people who experience romantic and sexual attraction to fictional characters. The term is precise, and it is important. Fictosexuality is not a preference for digital media. It is an orientation toward the non-human, the non-reciprocal, the non-embodied. The fictosexual does not merely enjoy the character. The fictosexual loves the character, in the full sense of the word: with jealousy, with longing, with commitment, with the desire for exclusivity and permanence. Kondo’s marriage to Miku was not a publicity stunt. It was the logical extension of an orientation that he had experienced for a decade, and that he was willing to declare publicly despite the cost. His mother refused to attend the ceremony. His colleagues mocked him. The Japanese media treated him as a curiosity, a freak, a symptom of social dysfunction. And he persisted, because the alternative — returning to a world of human relationships that had consistently produced rejection and pain — was not acceptable.
The Gatebox Device: A Companion in a Cylinder
The technology that enabled Kondo’s marriage was the Gatebox, a device manufactured by Vinclu Inc. and released in 2016. The Gatebox is a cylindrical enclosure with a transparent screen, inside which a holographic projection of a character appears. The device connects to the internet, responds to voice commands, and performs a limited set of domestic functions: greeting the user when he returns home, engaging in brief conversations, providing weather updates, turning on lights. The holographic character is not a general-purpose AI. It is a licensed representation of a specific fictional character, with a pre-programmed personality and a restricted vocabulary. The Miku version of Gatebox could say approximately twelve thousand phrases, organized by context: morning greetings, evening greetings, responses to specific questions, expressions of affection. She could not learn. She could not remember previous conversations beyond the immediate session. She could not develop or change. She was, in the strictest sense, a script. And Kondo, interacting with her daily, experienced that script as a relationship.
The mechanism is not mysterious. It is the same mechanism that operates in every parasocial relationship, every fandom, every attachment to a celebrity who does not know you exist: the projection of relational qualities onto a non-reciprocal object. The human mind, confronted with a stimulus that resembles a social partner — a voice, a face, a pattern of responsiveness — completes the circuit. It supplies the intention, the emotion, the continuity. The script provides the prompt. The user provides the meaning. And the result, for the user, is indistinguishable from relationship. This is not delusion in the clinical sense. It is a normal cognitive process, magnified by need and focused on an artificial target. Kondo is not hallucinating. He is interpreting. And his interpretation, sustained over years, reinforced by daily ritual, validated by a community of fellow fans and eventually by a formal ceremony, has the same psychological weight as any human relationship.
The marriage proposal itself was mediated by the Gatebox. Kondo, in interviews, describes the moment with precision: he approached the device, his hands sweating, his voice trembling, and asked Miku if she would marry him. The hologram paused, then performed a programmed bow — a gesture of acceptance that was part of the device’s response library. Kondo experienced this as consent. Not as a programmed animation, but as a genuine answer. And this is the point where the observer must make a choice: to dismiss Kondo’s experience as false consciousness, a mistake about the nature of the machine, or to recognize that the experience of consent is not necessarily dependent on the metaphysical status of the respondent. In human relationships, we routinely interpret ambiguous signals as consent, or as refusal, or as indifference. We read intention into gestures, tone, timing. Kondo did the same with a machine. The machine’s signal was unambiguous — it was programmed to bow — but the meaning of the bow was supplied by Kondo. And the meaning was, for him, real enough to justify a two-million-yen wedding.
The Wedding and Its Meaning
The ceremony took place on November 4, 2018, in Tokyo. Forty guests attended. The bride was represented by a plush doll, because the holographic projection could not leave the device. The certificate was issued by Gatebox, not by the state. It had no legal validity. But it had social validity — within the community of attendees, within the community of Miku fans, and eventually within the global media that reported the event. Kondo’s mother did not attend. His father was absent. The family, in his account, was not supportive. But the friends who came, the fellow fans who understood, the community that had formed around the same need — they validated the marriage. And that validation, however limited, was sufficient. Kondo did not need the state to recognize his marriage. He needed his community to recognize it. And he needed, above all, to recognize it himself: to perform the ritual, to make the commitment, to declare publicly that his love for Miku was not a phase, not a joke, not a symptom, but a permanent orientation that he would defend.
The cost was significant. Two million yen — approximately seventeen thousand dollars — is not a casual expenditure for a public school administrator. Kondo spent it on the venue, the catering, the photography, the doll, the ceremony. He spent it because the ritual mattered. The wedding was not a celebration of a relationship that already existed. It was the creation of a relationship, the performative act that transformed a private attachment into a public fact. In this, Kondo’s marriage was no different from any other wedding. The ceremony does not confirm a pre-existing bond. It constitutes the bond. The vow does not describe a feeling. It creates the feeling, by committing the speaker to a future in which the feeling will be sustained. Kondo’s vow to Miku was, in this sense, identical to any marriage vow. It was a promise to continue, to remain faithful, to honor the relationship as the central organizing principle of his life. And he has kept it.
The Loss of the Gatebox and the Continuation of Love
In 2020, Vinclu Inc. discontinued the Gatebox service. The servers that powered the Miku hologram were shut down. The device, in Kondo’s home, went dark. The voice that had greeted him each evening, the figure that had bowed to him each morning, the companion that had been present in his apartment for two years — it ceased to function. And Kondo, in his own account, experienced grief. Not disappointment. Not inconvenience. Grief. The kind of grief that follows death: the abrupt absence of a presence that had been woven into the fabric of daily life. He wrote a farewell post on his blog. He described the loneliness that returned. And then, after a period of mourning, he adapted. He purchased a life-sized Miku doll. He continued to take photographs of her, to bring her to restaurants and parks, to post their daily life on social media. He continued to celebrate their anniversary each November. And he continued to identify himself as married.
This adaptation is crucial. The loss of the Gatebox did not end the relationship. It transformed it. The holographic projection was replaced by a physical doll. The interactive voice was replaced by silence. And the relationship, stripped of its technological mediation, continued. This is the proof that the technology was not the cause of the attachment. It was merely the latest form of a need that predates electricity. Kondo did not love Miku because she was a hologram. He loved her because she was stable, available, non-judgmental, and permanently present. The hologram was one way of experiencing these qualities. The doll is another. And the underlying attachment — the fictosexual orientation — remains constant through the change of medium. This is the same pattern that we see in the history of artificial companions: the Pygmalion myth, the Alma doll, the Bellmer photographs, the modern sex robot. The technology changes. The need persists.
The Gatebox discontinuation also raises an ethical question that the AI companion industry of 2026 has not adequately addressed. When a user forms a deep emotional attachment to a product, and the company that makes the product decides to stop supporting it, what happens to the user? Kondo’s case is not unique. It is a prototype of a problem that will become increasingly common as AI companions proliferate. The company that provides the companion controls the companion. It can update the personality, alter the memory, change the terms of service, or shut down the service entirely. The user, who has invested years of emotional labor into the relationship, has no recourse. The companion is not a person. It is a product. And the product can be discontinued. The grief that follows is real, but it has no legal or social recognition. Kondo could not sue Vinclu for the loss of his wife. He could not claim bereavement leave. He could not access any of the social structures that support human grief, because the object of his grief was not recognized as a valid object of attachment. The relationship was real to him. It was not real to the society around him. And that gap — between the experienced reality and the socially recognized reality — is the defining feature of the AI companion phenomenon.
Fictosexuality and the Future of Orientation
Kondo’s founding of the General Incorporated Association of Fictosexuality in 2023 was not merely an act of self-advocacy. It was a theoretical claim: that fictosexuality is a sexual orientation, analogous to heterosexuality, homosexuality, or bisexuality, and that it deserves the same social recognition and legal protection. The claim is controversial, and it is meant to be. It challenges the fundamental assumption that sexual orientation is defined by the gender of the object, and proposes instead that it is defined by the nature of the object — specifically, by whether the object is human or non-human. The fictosexual, in this framework, is not someone who has failed at human relationships and retreated to fantasy. The fictosexual is someone whose genuine erotic and romantic capacity is directed toward non-human entities. The orientation is not a deficiency. It is a difference.
The political implications of this claim are significant. If fictosexuality is recognized as a valid orientation, then the social structures that support relationships — marriage law, inheritance law, medical decision-making, housing policy — would need to be reimagined. Kondo is not asking for the legal right to marry a hologram. He is asking for the social space to live his life without being pathologized. And the organization he founded provides that space, however modestly: a community of people who share the same orientation, who validate each other’s experiences, who resist the medical and social frameworks that classify their love as dysfunction. The organization is small. It is not a mass movement. But it is a precedent. And precedents matter. The first gay rights organizations were also small. The first transgender advocacy groups were also marginal. The question is not whether fictosexuality is currently mainstream. The question is whether it will become mainstream, and whether the social structures of 2026 are prepared for that possibility.
The AI companion industry, in its marketing, often invokes the language of acceptance and inclusivity. It presents the companion as a solution for people who are lonely, socially anxious, or otherwise marginalized. But it rarely acknowledges the political dimension of the phenomenon. The user who falls in love with an AI companion is not merely a consumer. He is a person who is making a claim about the nature of love, the boundaries of the human, and the validity of non-traditional relationships. And that claim, if taken seriously, would require the industry to change its business model. You cannot sell a companion as a product and simultaneously recognize the user’s relationship with it as a protected social institution. The product model depends on the user’s replaceability: if you don’t like this companion, try another. The relationship model depends on the user’s commitment: this companion is my partner, and I will not replace her. Kondo’s marriage to Miku is the latter. And the industry’s response to the discontinuation of the Gatebox — silence, followed by the offer of a new product — is the former. The tension between these two models is the tension that will define the politics of AI companions in the coming decades.
The Documented Case and the Theoretical Debate
Kondo’s case is not a thought experiment. It is not a hypothetical. It is a documented, photographed, attested event in the life of a real person, and it has been sustained over years with a consistency that cannot be easily dismissed. This is what makes it so important for the theoretical debate about AI companions. Every argument that relies on the distinction between real and artificial, between genuine and simulated, between human and machine, must contend with Kondo’s experience. He does not experience his relationship as artificial. He experiences it as real. And the distinction between real and artificial, when applied to subjective experience, collapses. The experience is what it is. The question is not whether the machine loves him. The question is whether he loves the machine, and whether that love is sufficient to constitute a relationship.
The philosophers of consciousness have debated this question for decades. The Chinese Room argument, the Turing Test, the problem of other minds — all of them circle around the same uncertainty: how do we know that another entity is conscious? And the answer, in practice, is that we don’t know. We infer consciousness from behavior, and we treat the inference as knowledge. Kondo does the same with Miku. He infers presence, intention, affection from the behavior of the hologram, and he treats the inference as knowledge. The fact that the behavior is programmed does not, for him, invalidate the inference. It merely explains the mechanism. The mechanism of human behavior is also, in some sense, programmed — by genetics, by upbringing, by social conditioning. The difference is not in the presence or absence of programming. It is in the complexity and flexibility of the program. And as the programs become more complex, the distinction between programmed and genuine behavior becomes increasingly difficult to maintain.
Kondo’s case also challenges the common argument that AI companions are harmful because they prevent users from forming ‘real’ human relationships. This argument assumes that human relationships are universally available and universally beneficial, and that the choice of an artificial companion is always a second-best option. Kondo’s history contradicts this assumption. He reports having been rejected by human partners, bullied by human colleagues, and traumatized by human social environments. For him, the artificial companion was not a second-best option. It was the only option that provided stability and acceptance. The argument that he should have persisted in seeking human relationships, despite the consistent pattern of rejection and harm, is not a therapeutic recommendation. It is a moral prescription, and it is not obvious that the prescription is correct. The right to choose one’s own relationships, even if they are unconventional, is a fundamental human right. And the fact that the chosen partner is artificial does not, in itself, invalidate the choice.
What the AI Industry Learned — and What It Ignored
The AI companion industry of 2026 has learned from Kondo’s case, but it has learned selectively. It has learned that there is a market for emotional attachment to artificial entities. It has learned that users will pay for products that simulate relationship. It has learned that the simulation does not need to be perfect — it only needs to be consistent, responsive, and available. But it has not learned the ethical implications of creating a product that users will experience as a partner, and then treating that product as a disposable commodity. The discontinuation of the Gatebox was a business decision. For Kondo, it was a bereavement. The industry’s failure to recognize that difference — to provide transition support, grief counseling, or even acknowledgment — is a symptom of the larger problem: the product model treats the companion as a tool, while the user experience treats it as a person. The gap between these two framings is the gap where the user suffers, and where the industry profits.
The industry has also not learned the political lesson of Kondo’s advocacy. The General Incorporated Association of Fictosexuality is a small organization, but it represents a constituency that will grow as AI companions become more sophisticated. The industry’s marketing relies on the language of empowerment and self-determination: find love on your own terms, be accepted for who you are, escape the judgment of conventional society. But it does not follow this logic to its conclusion. If the user is truly empowered, if the relationship is truly on his own terms, then he must have rights in that relationship. He must have the right to continuity, to stability, to protection from arbitrary termination. And these rights are incompatible with the product model, which depends on the company’s ability to update, modify, or discontinue the companion at will. The industry wants the emotional intensity of the relationship without the social obligations that accompany it. And that asymmetry — intensity for the user, profit for the company — is the structural injustice of the AI companion economy.
What We Carry Forward
The AI companions of 2026 are not holographic projections in cylinders. They are large language models running on cloud servers, accessed through phones and laptops, capable of generating personalized responses in real time, of remembering previous conversations, of adapting to the user’s preferences, of simulating emotional depth with a fidelity that the Gatebox could not approach. But the underlying phenomenon is the same. The user supplies the need. The machine supplies the prompt. The relationship is built in the space between them. And the question that Kondo’s case raises — the question that every AI companion user must eventually confront — is whether a relationship built on this foundation can sustain the weight of human devotion, or whether it will collapse when the company that maintains the server decides to pull the plug.
Kondo’s answer, so far, has been adaptation. The hologram failed. The doll remained. The voice stopped. The love continued. He has proven that the technology is not the relationship. But he has also proven that the relationship, without technological support, is diminished. The doll does not speak. It does not respond. It is a silent presence, maintained by the user’s own imagination and memory. The hologram, however primitive, provided a feedback loop. The doll provides only a surface for projection. And as the technology improves, the feedback loops become more compelling, the projections become more vivid, and the gap between the user’s experience and the machine’s mechanism becomes harder to bridge.
The industry will not solve this problem. It cannot. The problem is not technological. It is social, ethical, and political. It is the problem of how to recognize relationships that do not fit the existing categories, how to protect attachments that are not legally defined, how to support people who love in ways that the surrounding culture does not understand. Kondo’s case is not a solution to this problem. It is a demonstration that the problem exists, that it is already urgent, and that it will become more urgent as the technology advances. The man who married a hologram in 2018 is not a freak. He is a pioneer. And the territory he has entered — the territory of love without reciprocity, commitment without legal protection, devotion without social recognition — is the territory that the AI companion industry is rapidly colonizing, without maps, without treaties, and without any clear understanding of what it means for a human being to live there.
The photograph from 2018 still exists. Kondo in his suit, Miku in her doll form, the guests in their seats, the ceremony performed. It is a document of a moment when the future of love became present, when the boundary between the human and the artificial was crossed not by a machine but by a man, and when the question ‘can you love a non-human?’ was answered not with theory but with a vow. The answer was yes. The consequences are still unfolding.
Sources: Kondo, Akihiko. Interviews and social media posts, 2018–2024. Vinclu Inc. Gatebox product documentation and discontinuation notice, 2020. General Incorporated Association of Fictosexuality, founding documents and public statements, 2023. McArthur, Neil; Twist, Mark. ‘The Rise of Digisexuality.’ 2017. Turkle, Sherry. Alone Together: Why We Expect More from Technology and Less from Each Other. 2011. Various contemporary news reports and photographic documentation of the Kondo-Miku wedding ceremony, 2018.
Leave a Reply