The sheer scale of AI-generated media is daunting. Whether text, imagery, music, or even feature-length films, it feels as though the entire digital world is quickly filling up with artificially derived content. 

When we explore this landscape, we're hit by an equally vast range of emotions. In one moment, we're awed by the incredible realism and creativity, and in the next moment, fearful of the full implications of these innovations. But when we zoom out and consider the broad future of ChatGPT and Generative AI, one particular feeling may dominate: nausea. This isn't a nausea that's based in bodily physiology. Instead, it stems from a deeper, more existential place. 

It is the feeling described by the philosopher Jean-Paul Sartre in his appropriately titled novel, Nausea. 

He describes the character of Antoine Roquentin, who, while away on a writing project, slowly begins to lose connection with the world. He comes to feel alienated from the physical world, and ordinary items slowly lose their deeper meaning. Instead, they feel foreign and bizarre. 

This feeling of disconnect coincides with a looming sense of nausea. When he holds a knife, he is shocked by the raw sensation it produces in his fingers, as if feeling it for the first time. When he sees the root of a chestnut tree, he can't describe what it is or what it's for. Everything around him no longer has a deeper essence; it just exists. 

Nausea serves as a visceral metaphor for the feeling of alienation with the environment, and a complete lack of familiarity. As AI-Generated content begins to proliferate, we all may feel similar toward what we encounter. Like Roquentin, we may come to feel that the world - especially the digital world, is full of things that merely exist, devoid of their deeper meaning. 

How might this unfold? And what can the neuroscience of familiarity teach us about this “nauseating disconnect”? 

Let's dive in.

The Neuroscience of The Familiar

While the nausea that Roquentin experiences is metaphorical, it closely resembles the experiences of people with specific neuropsychological conditions. 

Gaining a sense of familiarity with things is ordinarily so seamless that we take it for granted. When an object catches your eye - a toy car, for example, your brain effortlessly retrieves its meaning: what it's called, what it's for, and how you feel about it. As we've seen, this includes the stories we've come to understand about an object that can instill a deeper essence. Familiar things automatically elicit a warm glow of acknowledgment. As soon as we recognize what we're seeing, our brains retrieve the object's underlying meaning. 

 

This process feels deceptively simple. But, under the hood, it requires careful synchrony of billions of neurons, primarily within the brain's temporal lobe. It's all effortless, automatic, and simultaneous - except when it's not. When damage disrupts this processing, we lose this sense of familiarity.  

Consider, for example, agnosia, a neuropsychological condition that typically results from damage to the temporal lobe. It's a selective deficit in retrieving meaning. The person can see the object perfectly fine. If you place the toy car in their laps, they describe its shape and colors without issue. They can feel its texture and describe the sensations. But they cannot tell you what it is, what it's called, or what it's for. There's no sense of familiarity. 

To the person with agnosia, the toy car isn't a "toy car"; it's just an object with specific sensory qualities. Like Roquentin's nausea, the thing is devoid of any deeper meaning. 

When All Familiarity is Lost: Cortard Delusion

A more extreme example of this loss of familiarity is Cortard's Delusion. Here, the patient experiences an extreme derealization: they can sense the world just fine, but nothing feels real. It's as if the world as a whole is devoid of meaning. 

The condition also resembles Capgras Delusion, in which people are convinced that the people in their lives have become imposters. Like agnosics, their sensory abilities are intact, and they can recognize them perfectly fine. But when they look at them, they don't sense the warm glow of familiarity they've come to expect. Thus, they appear to them as imposters. In Cortard Delusion, there's a similar loss of familiar glow, but now it's more widespread: gone is the familiarity with the entire world. 

The feeling is so extreme that the person becomes convinced they're dead. It's a bizarre conclusion, but like Capgras Delusion, it's the only one that "makes sense" when confronted with their strange experience of reality. If I'm alive, why do I feel absolutely nothing towards everything? Why is everything utterly unfamiliar?

In one case study, the patient WI was treated in the hospital for a head injury and later discharged to outpatient care. His sensory abilities were more or less intact, but, like Agnosic patients, he had difficulty recognizing people and everyday objects. His symptoms persisted and broadened. He claimed that nothing felt real and ultimately became convinced that he had died in the hospital and was somehow walking the earth like an empty corpse. 

Again, a seemingly bizarre conviction, but the only one that would make sense given the intense feeling of alienation from the world. 

Familiarity and Meaning in The Generative AI Landscape

Both Agnosia and Cortard Delusion are extreme examples of detachment and derealization. They're striking examples of what can happen when our brain's ability to create meaning and familiarity is disrupted. They illustrate that an experience similar to Roquentin's nausea, in which we feel alienated from our sensory experience, is not merely a literary device.  

At the same time, Roquentin provides us with another path to derealization. In contrast to these neuropsychological cases, he isn't suffering from any mental illness or neurological damage. Instead, he's undergoing a metaphysical shift that estranges himself from the rest of the world. But these issues aren't internal to him. Instead, the character's suffering and disillusionment come from the larger existential context. It's a disconnect that's as much about the broader circumstances as it is about his ability to connect with them. 

Like Roquentin, we're all faced with creating connection and meaning with the outside world. As Nausea helps illustrate, this is a universal challenge. Like Sartre’s ideas about freedom, seeking connection with the things in the world is part of human nature.

While this drive for meaning and familiarity is perennial, it's never been more challenging. The world is quickly filling up with exquisitely realistic AI-generated media. These replicas are entirely indistinguishable from what they purport to represent. On the one hand, these technologies have ushered in a new and exciting, creative world. But at the same time, they pose an unprecedented challenge to establishing that crucial sense of connection. 

Will we gain a foothold of deeper meaning in this new world of hyper-realistic artifice? Or will we, like Roquentin, be overwhelmed with detachment and nausea?

Check out Part 2 here, on How Generative AI Media Tweaks Our Psychological Schema

Photo by Matt Johnson via DALL-E

References for The Neuroscience and Nausea of Generative Artificial Intelligence

Sahoo, A., & Josephs, K. A. (2018). A neuropsychiatric analysis of the Cotard delusion. The Journal of neuropsychiatry and clinical neurosciences, 30(1), 58-65.

Sartre, J. P. (1938). Nausea. Penguin UK.

Squire, L. R., Stark, C. E., & Clark, R. E. (2004). The medial temporal lobe. Annu. Rev. Neurosci., 27, 279-306.

Young, A. W., & Leafhead, K. M. (1996). Betwixt life and death: Case studies of the Cotard delusion. Method in madness: Case studies in cognitive neuropsychiatry, 147-171.