Refik Anadol

Episode 10 C-19 Virtual Summit

Refik Anadol is a media artist and director originally from Istanbul, Turkey. He currently lives and works in Los Angeles, California. In this episode, Refik talks about his first projects, his connection and first approaches to AI and the future of machine learning.

I know you’re from Istanbul, living in LA; how did you end up there? How was this a breaking point for you?

I think the very first moment I knew Los Angeles could be a new home was when I was eight years old. A very cliché story: I watched the movie Blade Runner and I remember super clearly that the city of Los Angeles, in my mind, in my subconscious, was something that held the future: flying cars, the architecture, and where the cinema is reinvented. It's not only the tech world, but it's also close to Silicon Valley. It's in America, in a new continent that is different than where I lived; all these ideas were there.

I got my first master of fine arts in Istanbul in visual communication design. In 2011, I made a piece called Augmented Structures with a friend who was an architect, and I realized that as a solitary media artist I had very limited knowledge. In just three weeks we built a massive project. And that night, the opening night—again, this was nine years ago—we used sound data off the street and constructed a three-dimensional surface and projection method. No sleep; I just felt weird. The opening was amazing and people enjoyed it, but I felt like that wasn’t all the work I wanted to do in my life. It was amazing, but also kind of horrible! Then, I thought it was time to go somewhere else.

I openly looked at options in Europe and Asia, but I felt very connected to my school, which is UCLA, the Design Media Arts Department, where I got my degree. Some of my personal heroes work there, and, I have to say, I always dedicate my talks to my heroes and teachers. Without them, we are nothing. I had several great mentors like Casey Reas, Christian Molard, Jennifer Sinochem, and Rebecca Mendez. These are incredible people: media artists who have been using technology imagination for a long time. I said: "I'm going to learn from the best and then create a new journey with my studio (hopefully) when I'm done". That was the plan. In November of 2011, I set it up and now I'm in Los Angeles working with my team. It was a journey of eight years, but it's exciting. I’m so happy to be here in the city.


When did you realize you wanted to incorporate AI into your digital work?

That’s a fantastic question! I think I can start with the influence that the movie Blade Runner had on me. Blade Runner, for those who don’t know, is a story about replicants where a machine has this human-centric context of life. The city it inhabits always represented AI for me; from the moment I watched it, AI never left my mind.


I think my true connection with the idea of a machine was at eight years old. I got my first Commodore 128 and started playing games, I was an epic game addict for a long time! I was playing games too much, I guess. In games, when you think about an alternative opponent, a machine can also be your friend and play with you, beat you, or join you to compete with another machine. I like this feeling about AI; it is always on the menu. Like in Starcraft, right? So I had a question: "How can I purposefully merge AI into my artwork?" It began around 2012 when I founded my design media studio with Kayseri. We were discussing if data could become a pigment or data paintings then, with Christian Müller, we asked ourselves: can data become a sculpture? And we made the sculptures; those were the early steps in 2012 and 2013.


Around 2015 I was inspired by this amazing image that appeared on Reddit, forums, and everywhere. The DeepDream. For the people who don't know, it's basically from a research developed during the ImageNet Large Scale Visual Recognition Challenge in 2014 that uses a neural and coalitional network called Inception. The wonderful engineers at Google were exploring and experimenting with backpropagation, which is a technique that enables a machine to reinterpret what it learned by reconstructing it, based on what it knows. The idea that a machine can learn and interpret reality is so close to who we are as humans and to our human cognitive capacity of imagining, dreaming and hallucinating. And then, in February 2016 in San Francisco, a group from Google opened up an algorithm for anybody to use called DeepDream. Mario Klingemann and Memo Akten, who were two of the first artists working with this medium, did an auction with pieces created in DeepDream. It was an incredible revolution that an algorithm, an AI algorithm, was made public and that we could all use it. It was amazing. And that year I said, all right, this is it, it is time for me to learn AI. But the most amazing thing happened when I received an invitation to collaborate with Google's Artist and Machine Intelligence Group. It is a residency program inside Google founded by Blaise Agüera y Arcas, who is a wonderful mind, a brilliant engineer and a distinguished researcher. He and Kenric McDowell, who is the AI creator of the program, opened up this amazing opportunity for me. That led me to work with Mike Tyka in a team for almost a year. And that's how I started my journey with AI. I think I may be one of the earliest generations of media artists using AI and fortunately for me that was a big take off.


I was extremely inspired by the idea of consciousness and its context in AI. And we were very fortunate to have this amazing book by Jorge Luis Borges called the Library of Babel, which is an incredible book that tells the story of a library that contains all the information available in the world. And we started there. So the base idea was: What would happen if a whole archive could dream? And then we looked at the patterns of information produced from millions of images from a public library and let the archive dream its own contexts. And I became obsessed with the concept. For those who are not aware of this, when a machine learns, it holds the data in dimensions, a mathematical space, which is called Latent Space. And I was fascinated with this idea, like, how can we navigate in a world that doesn't exist? I couldn't even touch it. So I created all the stories and all these algorithms like lower dimensional reduction journey data, universe concept, and at the end this whole thing became an immersive piece. And now it is a studio that we have had for the last four years. We only focus on AI and its importance in the arts.


If computers now are able to “dream,” how do you envision the future of AI? Before you were talking about Blade Runner and how they imagined AI would be in the future, but today you are able to conceive a new take on AI. How do you see the evolution of AI in about 10 years?

Well, I have many heroes and sources of inspiration from science fiction. So, I always say that I get inspired from remembering the future. And you know I'm an artist. I'm a computer scientist. I am not like a neuroscientist. But I think, as a human being, I enjoy being able to create imagination. And I think, as Carl Sagan said, imagination will often carry us to the worlds that never were, but without it we go nowhere. I think it's an incredible statement from a scientist that has inspired many generations. So imagination is one way of looking at this question. We are still in the very first steps of it, but DeepMind, among others, is conducting incredible research. I was very honored to get a beautiful email yesterday from Demis, the CEO of DeepMind, because above all, my inspiration comes from the people who are doing this kind of research in an ethical, profound and purposeful way. So I think, as an artist, I am in a very unique position and have the advantage of being able to work with Intel, IBM, Microsoft, Nvidia because these are the people inventing our future and these are the people who are actually making new policies around the subject. So I feel like the only way to predict the future is by being able to work with these minds and understand their own concept of the future.


So what I can say is that everyone I am working and collaborating with are extremely positive people and they are not here to create an evil entity. They are not here to create Skynet. They are not here to make the evil algorithm that is going to destroy all known things or to make us lose our jobs. So first of all, let's get rid of that idea. I am very aware of human nature and I’m not ignoring those ramifications, but I do think we have to ask other questions.


For me the future of AI is about enhancing the power of the mind, expanding the power of the mind. It's not just cool technology with shiny pixels. It's truly about expanding our cognitive capacity. And when I think about that, I'm not thinking about the Elon style, I mean, these days Elon Musk is revealing a Neuralink demo, which will probably be a very evil idea, not for him to use of course, but maybe for someone else. By the way, he may be the brightest person in the world, but if someone opens the door to the human brain, it will be, undoubtedly, messed with. So let's work on accepting that fact. But what I'm saying is that when AI is purposefully used, it could help us find the cure for Alzheimer or cancer, or to find a solution to climate change, and find ways to improve learning. And I’m an artist, art is one of the means to bring awareness and enhance our imagination. But practically thinking, we have a huge chance of doing something that has never been done before for humanity.


I’m connecting your work because I know of your personal experience with your uncle with Alzheimer, so art is somehow reshaping the future, not only technology, there must be art, art seems to be experimental y expansive so if you look closely, you as an artist and as a creator are the one reshaping the way we conceive the future.

Yeah, so in 2016 I had the opportunity to witness the power of AI thanks to the Google arts residency, and the next year I was so unfortunate to see my uncle losing his memories to Alzheimer. For me, art is the only way to truly understand who we are and find that common language that connects us all. I mean, I don’t believe that a self-centered experience is purposeful or impactful enough to shape the future of humanity, but then again, it is art and we do have freedom of imagination. What really made an impact on me was the realization that memories are our most precious information, it is the most important data for our privacy and free will, I mean, for our future and our past. So I was super shocked by the power of this disease called Alzheimer's.


And it was just incredible to learn that you can actually look at the pattern of remembering and that the data that we as human beings are pulsing through our electric signals can be recognized and measured with AI. So, again, Artificial Intelligence is not about being flashy and fancy and cool, I sometimes feel like the field we are in is not conveying the right context and discourse. It’s a great topic to cover.


So, in my opinion, this kind of imagination, like Carl Sagan said, is where we can envision a future. So it's not solving the problem of Alzheimer's, but allowing the arts to create awareness that will hopefully help people appreciate the importance of remembering and being present in the world. At least this is my way of understanding this, again, it’s about expanding the power of the mind.


Do you see AI as a co-creator?

Absolutely! For me it’s like a thinking brush, a thinking pen. It’s allowing me to go beyond my own possibilities, again, it allows me to expand the power of my mind, that’s how I read it. It’s not something that can think or decide for me yet, but one day, maybe. And I mean, why not? It would be like having a colleague imagining with me, I would be happy to host that entity, but that’s not happening at the moment, not in this decade. To really understand this technology I believe we have to speculate, like architecture, which is not only concrete, glass or steel, architects are constantly speculating about the future, you know, life on Mars, life on other planets. They don’t build it, but they speculate and imagine it. It’s the same with AI, which for me is like a brush or a pen that can think and painting with a thinking brush is exactly the idea.


Can you share with us what is the creative process for creating new projects?

First of all, again, in 2011, when I started working by myself, I was too limited and I found out that being alone in this game is just not right. It's just not for me, at least. I'm an artist in the studio working alone with like a bunch of computers. It just felt too lonely. Since I was young I have always enjoyed hanging out with friends, you know, discovering new places or new objects to play with and just being curious together. I never lost that, I guess. And then I witnessed that when you work as a team you are more than one mind, you kill your ego and become this beautiful organic machine--which is how I refer to us “an organic machine”--you become something way bigger than who you are as an individual. I truly understood this when I saw our creations in 2014 when we were able to augment the Disney Hall, this incredible building designed by Frank Gehry. I was just so happy because this small team managed to augment Frank Gehry’s building; remember, six years ago using real time graphics and augmenting architecture was something very new and fresh. This was the first time I was doing something so public, I mean they would have three thousand people every night and let me tell you, Frank Gehry is tough, he doesn’t like copycats, he completely dislikes non-original ideas. But when he was in the green room, he was so supportive and his support was so meaningful that you feel an amazing energy and motivation. It’s your hero saying, OK, architecture, media, arts, they can collide and you don't need screens anymore.


Anyway, our first studio was in downtown Silverlake, in Los Angeles. We were like five people in the beginning and now we are 12. We speak several different languages. Twenty-five average age. Pretty much everyone can code in their own reality and we just try to be multinational, multicultural, do our best to be diverse and as equal as possible. We are collaborating right now with Google, DeepMind, Nvidia, Gensler, Epson, like these are big companies and they have incredible resources. So we are not alone, we have collaborated with these people for the last six years. Our model of imagination here at the studio is not individual, we always work as a collective, and it is not commercial, we do not do logos or advertising.


Here’s a question I always ask our guests. When do you decide that your art-work is ready? Because from your work I can tell that you are an extremely detailed oriented artist, so how do you decide when it is ready?

First of all, working with generative software is like a huge trap because once you are rendering a scene or rendering a moment, you have made your decision. You are saying, that's it. it’s done. That's an image. That's a video. But when you work with a living system like AI you don't have a beginning and an end, you are working with a system that is alive. When you are working with AI, in this latent space, which is infinite, you have this huge feeling of responsibility. Sometimes it's very depressive, I’ll tell you. Sometimes you start wondering, is this the right space? Is this the right image? Again, we are working with a thinking brush. So it's not like a mouse or a keyboard or some random numbers that run an algorithm. You have to ask yourself what is the part that you control, when do you become a part of the machine creatively, when do you become part of those numbers, how can you iterate that world that is beyond just machine randomness. So, I think that part is truly artistic. For example, our nature inspired pieces. I have always been obsessed with water, since I was a child, we are nothing without it and it’s ubiquitous so I find it as the most inspiring thing ever. I have been researching fluid dynamics since 2012 and that is not just a plugin or just numbers. It’s about the idea behind it. I said, OK, what would happen if the same data was looked at from the perspective of AI? What would happen if AI simulated life? That is nature. And that is how this whole series started, because when you start these ideations and you question the machine’s capacity of creating nature, then you have a whole new universe. You are not anymore plotting things randomly. It's not about creating simulations randomly and flip domains. It's not. So of course, you can stimulate whatever you want in the computer. But the real question is: why are you simulating rich data from a rich perspective? What is your point of view? I think sometimes that’s the point we are missing.

up