Explorations of AI Art — Episode 09
“An algorithm is just a chain of instructions that a computer runs to perform a task. In itself, it is just pure cold logic… However, when an artist designs an algorithm to express his creativity, the purpose of the code now serves as a source of emotions.” — Kirell Benzi
The New Year offers Cueva Gallery a journey into the artistic practice of Kirell Benzi, PhD in Data Science at EPFL (Ecole Polytechnique Fédérale de Lausanne), data artist, TEDx speaker and data visualization lecturer.
Before going into the maze of this type of art, however, for those who do not know what Data Art is, a short definition from Wikipedia is added here:
Information art, which is also known as informatism or data art, is emerging artforms that are inspired by and principally incorporate data, computer science, information technology, artificial intelligence, and related data-driven fields. The information revolution has resulted in over-abundant data that are critical in a wide range of areas, from the Internet to healthcare systems. Related to conceptual art, electronic art and new media art, informatism considers this new technological, economical, and cultural paradigm shift, such that artworks may provide social commentaries, synthesize multiple disciplines, and develop new aesthetics. 
Benzi’s work is displayed according to the nature of the data and guided by scientific rigor: his art is a balance of shapes and colors that underlie a deeper meaning and tell a story beyond human imagination. Distribution among nodes and edges, and relations between nodes in a network are represented by shapes, dots, lines where the purely creative part is in the choice of the color scale and the resulting shape of diverse combinations between networks. His work has attracted a lot of interest and has been shown in museums (i.e., ArtLab, Swiss National Museum), newspapers and magazines (i.e., La Recherche, 20minutes, Le Temps) and on over 100 websites in 10 languages (i.e., including Gizmodo, Engadget, The MarketPlace, TechRadar, Co.Design, Phys.org, VICE or Digital Trends). 
Kirell Benzi’s Data Art unveils the fact that calculation and data are not only cold, but can show a side able to generate curiosity and emotions and talk both to our mind and our heart. In other words, data have a “soul”.
Let’ s discover the “how” and “why” in our chat with the artist.
Beth Jochim: Kirell, could you tell us a little about your background and how did you become interested in connecting computer science with art?
Kirell Benzi: I was always fascinated by computers and started programming in my early teenage years. This passion led me to study computer science at the university and later drove me to get involved in Data Science with a Ph.D. from EPFL. I also started doing digital art at the same time with a software called TVPaint. I did not master it at all, but I loved the abstract shapes that I could create with it.
A few years ago, as I was trying to explain to people what my scientific research was about, and I realized that I could join my two passions and present science through art. It has been quite a journey ever since…
Beth J.: You belong to the movement of Data Art. Can you tell us what it is and how it fits in the realm of new media art?
K. B.: In one sentence, Data Art uses facts, information or statistics about a phenomenon to drive the artistic representation. It stems from data visualization and should be somewhat understandable by the audience from its description.
A cool definition that I found comes from the Data Art community on Reddit: “If you can hang a data visualization in your living room, it is probably data art”. It differs from generative art where the final representation often cannot be interpreted even though it might be driven by data.
Beth J.: Can you guide us into your creative process? How do you approach a new work and what are you looking for in a project?
K. B.: I start with the message and story that the future piece will convey. For instance, what are the different levels of understanding of the piece or what does it mean to create an artistic piece on this particular topic?
The second step is to look for a dataset that would be appropriate to answer all these questions. In bespoke pieces, I first help clients or collectors figure out how to get the most value from their data and choose the most impactful angle in terms of branding.
The third step is the data analysis part where I try to play around with the data to extract several interesting characteristics that will drive the piece. To be interesting to me, a project has to be visually complex and intriguing, I thus try to have as many dimensions as possible to choose from.
The final step consists in designing the visual with its colors. It is an iterative process where I try many different algorithms and techniques to get to a shape that I find interesting. The colors should also mean something in terms of interpretation by the audience and work with the overall message well.
Beth J.: Do algorithms have a soul? How and why?
K. B.: An algorithm is just a chain of instructions that a computer runs to perform a task. In itself, it is just pure cold logic… However, when an artist designs an algorithm to express his creativity, the purpose of the code now serves as a source of emotions. Poetically, we could say that he pours part of his soul into the code to connect with other human beings.
Beth J.: What are the difficulties that the Data Art movement is encountering? How could they be overcome?
K. B.: I think that the principal difficulty at the moment is that very few people are even aware it even exists. There is a noticeable difference between what the “art world” shows and how the general audience feels about it.
In the case of Data Art, I had an amazing response everywhere I went to present it. The idea of combining data and art, especially to communicate scientific ideas is gaining traction.
I think the solution could be to first reach out to a narrow audience of early-adopters and innovators. Hopefully, they will be in-turn evangelists and help spread the word.
Beth J.: Working on something new offers also the possibility to create interest and an audience that will follow you. How is your relationship with the public and with collectors?
K. B.: I am very fortunate to be able to present my work at numerous conferences around the world. I am currently preparing a “Data Art” tour in February 2020 with 12 dates in top US universities (Stanford, Berkeley, Princeton, Caltech, Cornell, etc.) to push its adoption even further. Collectors and museums are also getting more interested in the approach as our latest Jazz Luminaries project on the Montreux Jazz Festival archive showed.
Beth J.: What is a possible role for Data Art?
K. B.: A data artwork piece has all the elements to stimulate both our brain hemispheres with its emotional value and reflective aspect. Maybe I’m biased... but I think scientists and researchers could use more of this approach to explain their research. For collectors, it is also a chance to have a unique, personal piece of art created with something you deeply care about. More generally, it shows another more positive aspect of “big data” or data in general far from the usual scary representation depicted by the media.
Beth J.: Is there any question about art driven by science not asked yet and that you would like to answer?
K. B.: Funny question! Very recently we launched a scientific experiment to understand the impact of Data Art on people. We hope to show that art can also be studied scientifically.
If you feel like helping us figure out what makes Data Art special, please give us your honest opinion and evaluate some Data Artworks. It only takes about 5 min but the results would be a great source of insights to push this artistic field forward.
“I found I could say things with color and shapes that I couldn’t say any other way — things I had no words for.” — Georgia O’Keeffe
The combination of art and science is not new to Cueva Gallery’s readers. In the last year many artists have expressed a strong interest in the intertwining of these two worlds, only apparently opposed to each other.
This interview, however, has allowed us to know more closely some differences with other types of digital art. Like Benzi has explained, “Data Art uses facts, information or statistics about a phenomenon to drive the artistic representation. It stems from data visualization and should be somewhat understandable by the audience from its description.” This is one of the first differences with generative art, where the final work does not necessarily have a clear interpretation even if driven by data.
Another interesting point of reflection to me, again compared to Generative Art (where the discussion about who the artist is between the human and the machine is still ongoing) is the fact that the artistic approach described by Benzi does not allow any room for doubt. The scientist/artist has full control over the process, spanning from what story to tell to the level of understanding to offer the public, from what data have value to what level of impact they can bring. Colors and shapes are the artistic touch to a process that is highly guided by a scientific approach. But these are more than pure embellishments to cold data. They tell the other story, the one that has emotions and is cheerful and colorful: the story in which we can get lost and fantasize, hypnotized by light, hues and sinuous geometries.
Data Art gives us multidimensional storytelling that is powerful because it begins with the truth of data, and is then further enriched by the emotional side of the reality in which we live.
Thank you Kirell Benzi for this insightful chat! I have done my homework and participated in the scientific experiment you talked about above. I also would like to know more about what people think about Data Art, and what makes it so special. And I have more questions, so…
Let’s plan this for another time?
Resources and References