The Semiotic Approach to the AI Art of Fabrizio Augusto Poltronieri
Explorations of AI Art — Episode 19
"Philosophical unease provides the raw substance that I use to grow my artworks." -- Fabrizio Augusto Poltronieri.
[Fig. 2] Calliope (2014), part of the Iconic Theogonies' series; digital print 90x60 cm. Victoria & Albert Museum Collection. Credit: Fabrizio Augusto Poltronieri
What is your vision, as an artist and researcher, about the relationship between creativity and Artificial Intelligence?
I believe in a symbiotic relationship between artists and AI. This has been the main subject of my research and artistic practice for decades. I don’t think that computers are going to replace human creativity, but they will collaborate even more to enhance our perceptions and, of course, creativity.
This is not new, as the history of computation shows us that creativity is one of the fundamental foundations of computation. Nowadays, what can be considered new is the ubiquitous presence of computers and AI algorithms. It is exciting to see more people using computers as creative means. Still, this fact alone doesn’t imply that we are more creative just because we are surrounded by digital apparatuses.
As per today, even the most advanced AI system cannot do much alone. The human mind still plays a central role, and I don’t see it changing in the foreseeable future. As a semioticist, I do believe that there is a kind of mind in computers and that the real obstacle for us is to challenge it and accept to be challenged by it in creative and novel ways. This is what I mean by a symbiotic relationship: the human mind and the computer mind working, playing and creating together.
The recent advances in the field of Machine Learning are posing technical challenges and creating new issues. In the art field somebody would say that Artificial Intelligence is pushing us to rethink the very concepts of creativity and art. How is AI changing narratives and aesthetics?
To someone who is not aware of the history of art and technology this can be true. However, art and technology both share a common past that is reflected in the current days. What you are saying has happened before in history, and maybe the most remarkable example would be the Renascence period, when technological advances posed new challenges and created new issues to the field of art and vice-versa.
At that moment, the whole field of art has been questioned by developments in other areas, such as mathematics. Thus, to be honest, I am still waiting to see something impactful and creative made with the aid of machine learning that can be comparable to the ground-breaking art made by computer artists pioneers, like Frieder Nake and Harold Cohen.
This can sound a little bit bitter, but I think that the whole current Creative AI ecosystem is missing an appropriate aesthetic language engendered not by the eye, but by the mind. You see, Duchamp had already worked on this issue in the last century. He was also an enthusiast of the symbiotic relationship between machines and human creativity, advocating for a non-retinal art. So, the real issue for me is the problem of an original language of Creative AI, and this is the subject of my new book, due to be published next year, in collaboration with my colleague Professor Craig Vear.
[Fig. 3] Dionysus (2014), part of the Iconic Theogonies' series; digital print 90x60 cm. Victoria & Albert Museum Collection. Credit: Fabrizio Augusto Poltronieri
Part of your work, such as Iconic Theogonies, begins with the investigation of philosophical concepts and then with their transfer to artistic practice. Can you tell us about this approach?
All of my works begin with the investigation of philosophical concepts. As a mathematician and artist trained in semiotics, I am not interested only in the final results, but much more in the development process which, for me, always arises from philosophical questions I concern myself with. For example, I devoted one year to investigate how we visually represent and see ourselves through the lenses of AI algorithms. This was what I explored in my Selfie Apparatus series in which I used GANs.
I was not particularly interested in GANs per se, as for me they are trivial from an artistic perspective. If someone has some mathematical knowledge, access to a computer with a good GPU and has read some papers, it is possible for this person, by the current standards, to become an AI artist. However, I do not think that's enough. In my case, philosophical unease provides the raw substance that I use to grow my artworks.
In Pattern Recognition and Selfie Apparatus you create artworks combining symbolic AI and GANs. How do the old and the new come together in the artistic practice?
Due to my background as a programmer (I have been coding since I was eight years old), I found it much easier to tweak algorithms working directly with the code. Well, this is the basis of symbolic AI. I know that for the young people, sometimes it makes much more sense to grab a ready-to-go code from GitHub, throw some images or sounds in a folder, run the script and wait for the results. I like to do stuff the old way, understanding what is going on and writing code by myself, as I do believe that the act of writing code is a truly creative expression. It doesn’t mean that I dislike or do not use deep learning! On the contrary, nowadays I use neural networks all the time, but I like to combine the results I get from them with symbolic AI. Sometimes I also create my own neural networks models using TensorFlow, as I did with Pattern Recognition and Selfie Apparatus.
[Fig. 4] Selfie Apparatus 08 (2018), 30x30 cm., part of the Selfie Apparatus' series; digital print 30x30 cm. Credit: Fabrizio Augusto Poltronieri
[Fig.4] Pattern Recognition 05 (2019),part of the Pattern Recognition's series; digital print, 42x29 cm. Credit: Fabrizio Augusto Poltronieri
In your book Explorations in Art and Technology you describe "how artists have conceived and made new digital works from a historical perspective and how inter-disciplinary research has had a profound effect on the take up of digital technology in the wider community." Is the future of art an intimate collaboration between artists and technologists?
Yes, I think so. It is also important to highlight that I have been advocating for a long time the necessity of artists to learn how to code. This is urgent and pivotal if artists want to play any significant role in the future of humanity. The classes I run for the MA in Creative Technologies at the De Montfort University, in Leicester, England, are all about that: empowering artists through coding.
How do you envision the field of AI art progress over time?
My vision is a future where artists and AI work together in real-time, sharing transparent processes and decisions. The key concept for me is collaboration, and my current research efforts as an academic are entirely dedicated to that.
How are AI and digital media reshaping the relationship with the public?
I had a great experience about ten years ago, in my hometown São Paulo, Brazil, when I created from scratch, and using symbolic AI, an interactive robot whose main aim was to interact with humans trying to conquer their friendships. This robot was not a humanoid, and this fact made all the difference as the audience initially expected to see and interact with a human-like robot. After the initial surprise, in this major cybernetic art exhibition people could establish a stronger connection with this cylindrical shaped robot than with the other human-like robots.
From my point of view, if we want the public to engage with AI and digital media, we need to escape from common sense. We need to provide the audience with something that brings wonder (and I am not talking only about a wow effect), exploring new artistic formats that only digital technology can bring.
[Fig. 6] Selfie Apparatus 01 (2018), 40x40 cm., part of the Selfie Apparatus' series; digital print 30x30 cm. Credit: Fabrizio Augusto Poltronieri
Would you like to share with us a new project?
Currently, I am working on some Creative AI projects around music and on a big project to generate real-time urban landscapes from images I have been collecting in my travels around the world. Again, it is a combination of symbolic AI and neural networks, and it is going to be on display in a major international exhibition next year, connecting two cities in different continents.
References and Resources
About the author: Beth Jochim is the Creative AI Lead at Libre AI, and Director and Co-Founder at Cueva Gallery. She works at the intersection of technology and arts. She is actively involved in different activities that aim to democratize the field of Artificial Intelligence and Machine Learning, bringing the benefits of AI/ML to a larger audience. Connect whit Beth in LinkedIn or Twitter.
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.