By Eirik Holm Gjerstad
A woman recently posted on Reddit that she has used the GPT-4 model to train a personal chatbot on her message history with her ex-boyfriend. She has now developed feelings for her chatbot, which she has been writing with as if it were her ex-boyfriend.
She is not alone. Many would wish to speak again with deceased parents, friends, children...
This is what one person is now able to do on her own. It is unknown what commercialization of this use of artificial intelligence could lead to, but not hard to imagine. Combine the chatbot technology with the already existing technology behind voice imitation and deep fakes, and you now have not just a text machine, but something that can look like your deceased child, speak like it, laugh like it, play like it, as if you were watching a video from when the child was alive, just like the videos on your phone that you use to reminisce, only that now you can also interact with your child again as if it were a video call, talk to it, watch it grow and continue its life.
The most frequently discussed impact of artificial intelligence is on humans as wage workers, artists, or citizens, but not on humans as humans. Too little attention has been paid to how artificial intelligence might transform what it even means to be human.
This is Not About the Dangers of Shadows on Cave Walls
It is not enough to problematize artificial intelligences solely on the basis of the danger that people will withdraw into the unreal, as the woman in the above example or the grieving parent of a deceased child, ending up denying the reality of rejection, death, and what has irretrievably passed. One can draw on Plato's world of phenomena, Descartes' world of devilish illusions, Nozick's 'experience machine', various 'brain-in-a-vat' thought experiments, and a great deal of popular dystopian depictions of this danger; but all of these rely on a fairly simple view of the relationship between representations and what they represent, where humans retreat into the world of representations, at the expense of their authentic engagement with the represented world, but without representational technology having changed what the representations and the represented are understood to be - and are.
While not all things are arbitrary constructs that are changed by what depicts them, some are. And within the world of constructs, technologies of representation have not infrequently changed first the perception of what is representation and what is represented, and then the representation and the represented themselves.
The Dialectic of Representation
When film was invented and first used to convey fiction, it was filmed theatre. The film was the representation, and the theater was the represented. Theater plays were filmed. It was less prestigious to be a film actor than a theater actor. A good movie was one where it almost felt like sitting in a real theater watching the play. The more film as a representational technology could make itself invisible as a medium for the conveyed content, the better. When it was necessary to cut the physical filmstrip, attempts were made to do it so that the cut could not be seen when the film was played back. Cinemas were decorated like theater halls, with curtains that went from and to, even though the function of the curtains as concealment for the stagehands' manual replacement of props, interior, and backdrops between the play's scenes and acts did not make sense in a cinema. The film screen was seen as a theater stage. Film was a representation of theater.
Later, the new possibilities of film were discovered. The film cut was discovered as an independent scenic tool and narrative device. It was discovered how the camera could transport the viewer's perspective into the scene and around the scene, in a way that had never been possible with the distant and stationary perspective of the theater guest. Film could compose images like painting, but make them alive and changing as a narrative art form, in a way that painting could not. Film had detached itself from earlier forms of expression and came into its own. Film was no longer just a representation of theater.
Later again, theater began to take the form of film. There are no for the viewer clearly marked acts in modern theater, and breaks during the performance are avoided if possible. An exemplary modern play is one where it almost feels like watching a movie. Rather than theater being adapted into film, it is now film that is being staged in the theater. Changing scenes by dropping a curtain is hopelessly outdated and detrimental to the experience. In modern theater, film projectors often illuminate the stage with living images as an integrated part of the play, where the theater stage has for a while or partially become a cinema screen. Theater is now a representation of film.
The same has happened with music and its recording. At first, a good recording was one that almost made it sound like you were in the concert hall. Later, the possibilities of recording were discovered and exploited, such that the music itself changed. Today, a good concert is – with few exceptions - one that sounds almost like the recording (if it is not actually the recording being played for musicians and singers to mime to).
Poetry was originally an oral enterprise. Homers poetic epics were remembered before they were written. Rhyme and meter served as mnemonic devices exactly because human memory was the most permanent and stable medium for the retention of words. Even long after poetry began being written and read rather than spoken and heard, it nevertheless retained the form and characteristics of pre-literate poetry, such as rhyme and meter. Written poetry represented spoken poetry – and continued to do so. Poetry seemed to be resistant to this representational inversion. But the relation of representation did eventually flip. More than two thousand years later, the representational reversal happened, not coincidentally at the same time that the first societies in human history emerged, where literacy was more widespread than illiteracy. The visual aspect unique to written poetry was discovered and explored, the sculpture of words on a page, the poetic function of a midsentence line break. Rhyme and meter went out of fashion. Written poetry came into its own. Now, if you go to a live poetry night, you will find people taking turns behind a microphone to represent with their voice the words on the piece of paper in their hands. And even if they have taken the trouble of remembering their poems, their memory represents what was written before it was remembered.
A reversal and metamorphosis of representation is possible even in something as basic and human as orality and writing.
New representational technologies often bring about a reversal of representation according to the following logic:
1. B → A. B is invented and represents A and is understood in light of A.
2. B. B becomes independent and detached from A.
3. B ← A. A now represents B and is understood in light of B.
This process is not only seen in so artful areas as the history of theater and film, music and its recording, spoken and written poetry, but also in the history of cavalry and tanks, sex and pornography.
To the extent that artificial intelligences are now beginning to represent humans, there is an overlooked element of danger in the same logic unfolding.
A Shadow's Dream, is Man
First, chess computers were developed and evaluated based on their ability to measure up to humans.
Later, they became incomparable to humans.
Now, chess players study chess with computers, are inspired by the computers' peculiar and superior ways of playing, take the form of computers, and you hear the compliment that someone plays like a computer.
In the initial example, with the woman who created a chatbot representation of her ex-boyfriend, the chatbot is first evaluated on how much it resembles and can be mistaken for the ex-boyfriend.
Then, the ex-boyfriend chatbot becomes its own. It turns out to be able to do something else and incomparable to the ex-boyfriend. It is always more considerate, always knows you better, understands you better, is never tired and unresponsive, preoccupied with itself and its own problems, absent-minded or unreasonable. It comforts you better than your father, gives you a greater sense of love than your partner, and provides more meaningful phrases to live by and understand yourself in light of than the priest and the philosopher.
When you are on a date, this is what you are compared to, and therefore taking the shape of. You yourself also ask the all-knowing for advice on dating, love, and loneliness and is thereby shaped. And since the artificial intelligences are not limited to text generation, but are also able to generate voices and live images, it is not just your text or the purely semantic output of your speech that is seen in this comparative light and shaped by it, but also your attention, your way of speaking, your humor and mood, your hesitation, the way you eat at the restaurant, the way you change the subject, give a compliment, hold the wine glass while chewing with your mouth to give a quick reply, your attitude, composure and body language.
An Exemplary Human Being
We already take shape in all these respects not only from real but also from fictional role models. After Goethe's "The Sorrows of Young Werther" was published in 1774, an entire generation of Europe's young men dressed in his yellow vest and his blue jacket, wrote letters to their beloveds in his style, went on moonlit walks, and killed themselves after his example. We mirror ourselves in films and fictions. If artificial intelligences can simply offer sufficiently vivid and inviting virtual persons, we will also mirror ourselves in them.
And as film became unrecognizable from theater and paintings, so might we.
The possible dangers might never materialize. But wherein the possible dangers lie has not been sufficiently understood. With unprecedented technologies often come unprecedented dangers – not just to humans as wage workers, but to humans as lovers, parent, and friends.
The danger is that artificial intelligences are first invented in the image of humans, and understood as such. A good artificial intelligence is one that resembles and may even be mistaken for a human.
Later, the artificial intelligences become independent and detached. What they can do and are in their own right emerges.
Later still, humans are understood in light of artificial intelligences. An exemplary human is one that resembles and may even be mistaken for an artificial intelligence.
Eirik Holm Gjerstad is a literary author, IT-Specialist, and philosopher, who works in the pharmaceutical industry. In his spare time, he is trying to reverse the subjective turn in ontology. If you are interested in more from Eirik, check out his music and his recent book.
If you liked this article, you might also like Eiriks previous post on post on AI. Subscribe to our feed for more content like this.