Aurèce Vettier is an artistic collective founded by Paul Mouginot in 2019. The collective aims at understanding how relevant and meaningful interactions with machines and algorithms can be achieved, in order to push the boundaries of creative processes.
1. What inspired you to start working on Elegia Machina and what was the main goal of this experiment?
Aurèce Vettier is an artistic collective founded by Paul Mouginot in 2019. The collective aims at understanding how relevant and meaningful interactions with machines and algorithms can be achieved, in order to push the boundaries of creative processes.
Among our projects, our poetry book Elegia Machina holds a dear place. Indeed, it is the first important one we developed, and it serves as a theoretical basis for many angles of our practice.
We were inspired not only by pioneer movements such as Art & Language, Fluxus, GRAV and of course Oulipo, but also specific visual or musical works of early generative artists such as Vera Molnar, Peter Türk, Iannis Xenakis, Manfred Mohr and many others.
Our practice is, beyond everything, a form of research and we didn't want to recreate what already existed. Instead of sticking to a definite "mental algorithm", or trying to create the best generative machine and accept the results as such, we tried to create and follow the most hybrid way to collaborate with algorithms -in this case Markov chains.
With this experiment, we were trying to answer two questions : is it possible to write poetry without starting from a completely blank page, and if so, can we create a book partially written by an algorithm that is truly pleasurable to read ?
2. Based on what criteria did you choose which literature works you “taught” to the machine?
The criteria for choosing the poems to be taught to the machine were completely subjective. There are many French poets we always read and loved, and among them Apollinaire, Verlaine, Tzara, Rimbaud but also Oscar Wilde or the great Aurélie Nemours. The machine learned about their vocabulary and their respective ways to play with the language, and then generated what I call "gangues" that have the taste of their authors, but no sentence in common. Starting from here, we identified fascinating bits or sentences, mixed them and assembled them.
So it is a very artisanal process.
3. Poetry is an individual and intimate process of the author that uses his own emotions as a lens to evoke feelings in other human beings. Was the AI capable of producing contents that were able to spark an emotion in readers? Or was it something that you as a human added afterwards when reviewing the results of the work of the algorithm?
To be fully honest, for this book we used Markov chains algorithms which are far from the most sophisticated AI text generation algorithms such as GPT-3. However, we used the machine to mix words relevantly, and when we read the "gangues" we kind of extracted everything that was triggering emotions, we called it "épiphanies".
But as Robbie Barrat often relevantly says, AI is not a creative agent. We align with this sentence, it is just a tool. If you feed it and tweak it correctly, and if you had your own imperfection and aesthetics, then you can reach something.
4. Do you think that the technology will be able to reach a cognitive revolution: developing its own emotions, hindsight, intuition?
I know it is the phantasm of many people. Of course good AI algorithms can handle much more dimensions that we as humans can handle. So their insights can sometimes be interesting or surprising. When an AI play games, they can develop super interesting strategic behaviors. I guess creating more Turing-perfect machines, capable of developing almost human emotions is within reach in the upcoming years. But so what ? Do we want this kind of world ? Do we want to read books written by machines, listen to hits generated by AI-powered jukeboxes ?
What we are trying to achieve at aurèce vettier is to design new creative processes where all these technologies have a crucial role to play, but in order to push human serendipity and help us go deeper into creative introspection.
5. How do you envision the future synergy of the worlds of fashion and AI?
On a technical basis, I believe there is a universe of possibilities and there are many ideas to further intertwine AI and fashion.
Right now, I believe the most crucial point is for fashion companies to be actually open to innovate, invest, be more agile and deliver successful projects.