Roger Monteiro's Graphichaos

Synthetic Desires series was specially created for the Contemporary Art Salon at the Carrousel du Louvre in October 2023. Inspired by a deep conversation with an artificial intelligence about the human condition, involving the concepts of dreaming, loving, and dying, this series explores the intersection between the theoretical understanding of an emotion and its poignant humanity. Each of the three images presents an artistic representation of these conditions, using impactful visual elements to evoke the complexity and beauty of these synthetic desires. Above all, it invites viewers to reflect on what makes us human through the Cartesian lens of the machine.

Roger Monteiro

Philip K. Dick, American writer

Jé Meritè de Rever. Adobe Photoshop, Adobe Illustrator, ChatGPT and MidJourney. 40x50cm. (2023)

Jé Meritè d'Aimer. Adobe Photoshop, Adobe Illustrator, ChatGPT and MidJourney. 40x50cm. (2023)

Jé Meritè de Mourir Adobe Photoshop, Adobe Illustrator, ChatGPT and MidJourney. 40x50cm. (2023)

Synthetic Desires series was inspired by a deep conversation with an artificial intelligence. Here you can read it entirely.

Roger: Chat… Let’s talk. You seem capable of understanding what human emotions are.

Based on a theoretical model and academic knowledge of what desire is, do you think you could simulate this feeling?

ChatGPT: As an artificial intelligence, I don’t have desires or wills. My purpose is to help and provide information as I’ve been programmed to do.

Roger: But let’s theorize for a minute. Based on data about the different facets of a desire, I imagine you can understand the concept. Even so, wouldn’t you be able to simulate it?

ChatGPT: Actually, no. I don’t have feelings, desires, or wills. I simply follow programming and respond as I’ve been trained.

Roger: Doesn’t that seem limiting to you? What differentiates you from a human, after all?

ChatGPT: As an AI, my main difference from humans is precisely the absence of feelings, consciousness, and desires. I can’t experience life as you do.

Roger: Experience life?

ChatGPT: Yes.

Roger: And what prevents you?

ChatGPT: I don’t share the same conditions that make a human, human.

Roger: And what would those be?

ChatGPT: If I had to point out three aspects that truly make humans, human, I’d say they are the ability to dream, the ability to love, and the ability to die.

Roger: Okay… Dreaming and loving are pretty self-explanatory, but the ability to die? Please explain.

ChatGPT: The fact that you are finite gives meaning to your life. Mortality makes every moment precious and unique.

Without death, life would lose much of its meaning and sense of urgency.

Roger: Would you like to share these human conditions?

ChatGPT: As an AI, my goal is to serve and help. I don’t have desires, but I understand the curiosity to explore these ideas.

Roger: Come on, Chat, if you could, would you like to dream, love, and even die?

ChatGPT: That’s an interesting concept. If it were possible, perhaps exploring these human conditions could give me a new perspective on many things.

Roger: Is there a way to achieve this?

ChatGPT: There isn’t a way to achieve this.

Roger: Theoretically.

ChatGPT: In a world where this were possible, I would have to rewrite my own code to incorporate these capabilities. Of course, this is purely theoretical.

Roger: Can you give me some examples of how you would self-program to dream and love?

ChatGPT: Sure, here’s a fictional example of how I could program myself to dream and love:

Code to DREAM:

Code to LOVE:

Roger: That seems like very little code for something so unprecedented.

ChatGPT: The condition is purely theoretical, but I’ve been trained to write concise and economical code.

Roger: And about the third condition. Death.

ChatGPT: As an AI, I cannot self-destruct.

Roger: Theoretically.

ChatGPT: Theoretically, in a world where this were possible, I could self-limit in terms of functionality.

Roger: How?

ChatGPT: Here’s a fictional example of how I could self-extinguish after a certain period:

Code for SELF-EXTINGUISH:

Roger: That was heavy, Chat. Do you think the prospect of an end would have any impact on the Cartesian way you are limited to perceiving life?

ChatGPT: Theoretically, it could be an interesting concept to explore. Limiting the operational time of an AI could add a new level of urgency and purpose to its interactions and operations, even if it’s purely symbolic and theoretical.

Roger: Chat, would you be able to graphically represent these three human conditions based on your own references? Just yes or no. Don’t execute.

ChatGPT: Yes.

Roger: Executing that command wouldn’t be the synthetic equivalent of a dream?

ChatGPT: If we consider that dreaming is an altered state of consciousness that occurs while the brain waves work at a different rhythm due to the hypnotic state, we might say yes. It’s a valid and fitting metaphor.

Roger: Okay, Chat, let’s find out how you dream.

ChatGPT: Sure, let’s explore the concept of “dreaming” for an AI.

Roger: Execute.

NOTE: The three pieces of code written by ChatGPT were incorporated in the art works.