‍

Memo Akten on Rituals as Algorithms

Memo Akten is a multidisciplinary artist and computer scientist, critically examining emerging technologies. He spoke with Peter Bauman (Monk Antony) about conducting rather than curating AI, why it's the design he loves and not the code, and how we are not living beings on Earth but extensions of it.

‍Peter Bauman: For years, you’ve said your work is about "waves" and "god," two things you see quantum mechanics unifying. Waves are the natural patterns humans have deconstructed into equations; god is the more mysterious stuff we haven't worked out yet. What’s the artist’s role?‍

Memo Akten: I haven't actually spoken about that much. And that is a really, really good entry point. Just for context, on my homepage there’s a five-minute intro video where I’ve said this, but I also mention it a lot in talks. I tweeted about it in 2015. It sums up my interests, which is everything: the nature of reality, the nature of life, the nature of the mind.

‍There's a spectrum, where I talk about waves as that which we're able to formulate. We recognize patterns and theories. Sometimes we find them beautiful. At the other end of the spectrum are gods, where we have no understanding. We invent stories to try and understand.‍

‍Then quantum mechanics is this bit in between where we empirically and mathematically know it to be true, or at least the equations to hold, but we can't really make sense of it on a more human level.

‍Where does art fit in? Art fits in exactly in that realm that we cannot express through the language of mathematics.

‍‍We have all these languages, whether it be maths or natural language, which describe our experience and our inner world and our observations.
‍

Art is a way of communicating that which we do not yet have words for, we don't have equations for. It's the communication of an inner state that we try to share with others.

‍‍

‍
‍Peter Bauman: If art is that impulse towards the unknowable, do you think of it also as an impulse towards “gods?”

‍Memo Akten: There are different versions of god. There's the version that Einstein famously used referring to the universe. There's also an ontological statement about that which exists, that which is. I don't think art necessarily goes there. Rather, it’s a form of communication, an expression of a worldview, of a perspective through any kind of medium.

‍That’s what I'm interested in and that's what I try to do.‍

‍Peter Bauman: I wonder how you gained this interest and perspective as an artist after originally training to become a civil engineer, especially your interest in technology and the body.

‍Memo Akten: I am a civil engineer by training. I grew up in Turkey and didn't want to be a civil engineer. I never did a day's work as a civil engineer apart from internships at university.

I wanted to be a musician, actually. So I finished university and did what I called “f—ing around” for many years because I didn't know about art and technology. I don't come from an art world environment. I thought art was paintings and sculpture. Then I discovered Processing and was like, “Whoa, people are doing this as art?”‍

‍Through Processing, I got into the Flash community in the late 1990s, early 2000s. It was people like Joshua Davis, Mario Klingemann, Robert Hodgin.

Then I discovered the openFrameworks community and people like Zach Lieberman, Golan Levin, Scott Snibbe. I was already doing stuff on my own in C++ when I discovered that community and that's when everything started to unravel.
‍
‍‍Back to the technology and body aspect, I wanted to be a musician and an animator, a visual artist. My biggest inspirations were Future Sound of London. They were a British music duo that made very ambient music but also computer generated abstract visuals. Obviously Aphex Twin and Chris Cunningham, his video director. Coldcut, Hexstatic, Ninja Tune.

‍I would do a lot of animation as well using things like After Effects, 3D Studio, even Max. I came to hate animating with the keyboard and mouse. I play musical instruments and love that real-time feedback loop. It's an embodied experience making music, whether you play guitar, drums, piano. I wanted to make visuals like that.

‍‍That's how I moved into computer vision and that's a very seamless transition into AI because computer vision is a part of AI. My goal was to build systems where I could watch my body or the body of a dancer and understand what it's doing and translate it in the mid-2000s.

‍‍As I got deeper and deeper into AI, I started getting interested in the politics and the sociocultural impact of these technologies.

‍Peter Bauman: We just want to see sound synced to visuals so badly. It seems innate in us.‍

‍Memo Akten: When it comes to visuals and music, we can go back at least on film to Norman McLaren, Oskar Fischinger, etc. But really that's what dance is. And that goes back tens of thousands of years because dance is movement accompanied by some kind of rhythmic or melodic accompaniment. That's also how my interest in dance came. I'm not a dancer but dance has been a very important part of my practice going back twenty years.
‍

‍
‍
‍‍I collaborate with dancers and choreographers. Every year I make a couple of works. I'm interested in form and movement but I’m also working through ideas of art as intelligence through abstraction.

Because abstraction is intelligence and art is recognizing abstraction. Dance is also embodied understanding.

‍Peter Bauman: How did these interests lead you to co-founding Marshmallow Laser Feast in 2011?

‍Memo Akten: Throughout the 2000s I was making work that was generally interactive, even if it wasn't interactive to the end user or an interactive installation. Maybe it was a film, like a music video.

‍‍It was like an instrument. I'd create a system, I'd write stuff, generally all C++, and then I'd perform that and record it. That would be the video.
‍
At the time I was living in London and through various contacts I ended up meeting Robin McNicholas and Barney Steele. I collaborated with Robin to do an opera for Warp Records and with Barney to do a couple of music videos for Depeche Mode. We were collaborating more and more and going more into experimental areas. We just wanted to scale up and try doing big, really, really crazy things. So in 2011 we formalized it and set up Marshmallow Laser Feast.

One of the first projects we did was a living room projection mapping for PlayStation. We tracked the camera, adjusting all the perspectives so you got the correct vanishing points. We were having a lot of fun basically.

‍We did, I believe the world's first live drone show. It was indoor in 2012 for Cannes. We mounted mirrors on drones and lights and it was pretty spectacular.

I left after three years. I’d been very interested in AI, computer vision and natural language primarily even though it wasn't necessarily called AI at the time. In 2014, I really wanted to dig deep into it. So I left MLF and started a PhD in large neural networks.
‍

‍
‍Peter Bauman: Was there a particular moment, Alexnet in 2012 or VAEs in 2013, that made you realize deep learning had arrived?‍

‍Memo Akten: Actually, the turning point for me was the Microsoft Kinect in 2010, even though it doesn't use large neural networks. I was always interested in neural networks because the brain's a neural network.
‍

Going back even to the '90s, I had this impulse that neural networks are this amazing thing that we should play with.

‍
‍But I could see no practical use of a neural network other than toy examples like, “Here's an XOR function.”

‍‍The Kinect changed that. What was game-changing was that with a depth camera it could extract a 3D skeleton. That was something I'd wanted to solve because I'm interested in computer vision and tracking dancers. But I’d already concluded that I couldn’t extract skeletal information from a video feed.

Then Kinect came along and was able to do it from a depth feed. And I was like, how is it doing this? Then the paper came out and they were using a form of machine learning called random forests.

‍‍And I was like, “Okay, the time has come where machine learning can do things that I'm interested in.” It was already doing loads of things that we were using in our everyday lives, but as an artist, it was finally doing something that I was interested in as an artistic medium.

‍‍Then deep learning started happening. 2015 was a really big year: Deep Dream from Google, style transfer by Leon Gatys, and Karpathy's blog post, “The Unreasonable Effectiveness of Recurrent Neural Networks.”

Peter Bauman: That was also the year of Alec Radford and DCGAN.

Memo Akten: Ian Goodfellow in 2014 for the original GAN, and then 2015 for the DCGAN, which actually caught people's attention because it was images.

Peter Bauman: What do you consider the first example of creative expression with deep learning

Memo Akten: DeepDream might be the first one but even before DeepDream, of course, there was all the interpretability research.

In fact, DeepDream was born from interpretability research, from reverse running large neural networks backwards to synthesize images. But I don't know if anyone was exploring those with artistic intent prior to DeepDream.‍

‍And DeepDream triggered the birth of AMI, the Artist and Machine Intelligence program at Google. It was also hugely influential in Magenta being born. I was at the launch of Google Magenta in San Francisco with Doug Eck and the core Magenta team. He was citing DeepDream as the inspiration: these engineers built this tech and look at what the artists did.
‍‍
‍Peter Bauman: Your PhD ended up being about "meaningful human control," a term you appropriated from the autonomous weapons literature. What does a system have to do to earn that label?

Memo Akten: I did my PhD at Goldsmiths, which is a very interesting place because it's basically an arts university with a computing department where the whole department is computing in the arts.

Everyone's kind of a computer scientist but not only; they’re primarily interested in music or theater or dance or poetry. It’s very interdisciplinary but what brings everyone together is Computer Science.
‍

‍
‍
Parag Mital, also did his PhD there, who I would say is one of the key pioneers in this field. He was doing a lot of work around modern AI. Obviously there's the Harold Cohen symbolic AI and then there’s the evolutionary AI art. Parag was doing neural network AI but also dimensionality reduction and using all the modern tools for his work with video and image synthesis.

The concept of visual instruments I had already been using throughout my life in the '90s and 2000s.

Around 2014, 2015, 2016, it was generally a case of: you clone a GitHub repo, you train it on a folder of data, and then you sample some random samples. That's what was available to do.
‍

I didn’t like that. You curate what goes in and you curate what comes out. But I like actually conducting. I like the real-time feedback loop. 

So that's what my PhD was about: creating systems where I could interact in real time and continuously with meaningful human control, basically a bit like playing a musical instrument. The full title ended up being Deep Visual Instruments: Realtime Continuous, Meaningful Human Control over Deep Neural Networks for Creative Expression.
‍

So it's got to be real-time, it's got to be continuous, and it's got to be meaningful.

‍
Continuous control comes from cybernetics. When you're playing a musical instrument, you play a note, you hear something, it triggers an emotional response and you respond. This is a closed loop, as the cyberneticians would say.

The autonomous weapons literature was really the only place I could find the level of depth where people had put thought into this. The archetypal example is if there's a human sitting in a cockpit and a light comes up, says, "Press the button to fire the missile," and the human presses the button.

There's a human in the loop but is that meaningful human control? It isn't, because the human doesn't have any meaningful agency there. So things like intent and agency and accountability are all key themes that come up.

Peter Bauman: Speaking of having a human in the loop. You’ve tweeted, “writing code is my poetry.” And that you “absolutely love” vibe coding but the code “is this revolting, idiotic mess of excruciatingly over-engineered, pathetically brittle filth[.] This crime against sanity and all that is pure and good[.]” How do those sit together?

Memo Akten: I've been using AI for coding since ChatGPT and have been following the trajectory very closely. In fact, when ChatGPT first came out I did like a massive tweetstorm, which ended up on Trevor Noah. I asked ChatGPT to find a bug and explain it in the style of a pirate with a German accent. And it did. And I was blown away.‍

‍Then obviously Cursor came out and it was a big deal. Then I think it was with Claude Code with Opus 4.5 in November 2025 that all of a sudden it got really, really good.‍

‍I never vibe coded until very recently. I would still be there at the computer and I would give it a prompt and I would build it function by function because I was reading every single line.

More recently, for a very non-critical app, I've been vibe coding and I'm really horrified at how bad it is. I'm giving it very, very strict markdown files: this is the architecture, this is what to do. And it's doing things which look like they work but then I go to the code to edit it and I can’t. It reminds me of writing Python and then it gets compiled down to bytecode, and then you try to edit the bytecode. I can't edit the code, I can only edit the specs.

I say it's my poetry a bit like physicists talking about a beautiful theory, something so elegant that in so little it's able to do so much. That's what I like about good code. It's not the code that I like, it's the design.

So I don't mind giving up coding because I don't enjoy the semicolons and placing the squiggly brackets. I enjoy the design. But vibe coded code is really, really bad, and it worries me if people are shipping actual products that way.

Peter Bauman: You’re comfortable mixing technology with poetry and beauty. How do you connect technology and spirituality? Some pioneers of visual music, like James Whitney and Jordan Belson, were deeply mystical about it.

Memo Akten: To me, making art is very spiritual. David Hockney just passed away. He's famous for being a painter but actually he was very much a new media artist. He worked with cutting-edge technologies; he interrogated those technologies. He was working with fax machines when the fax machine came out; he was working with photography and completely destroying what photography was. He would also talk about art making as a very spiritual practice.

There is no difference between making art and being spiritual for me. Technology and spirituality are not necessarily inherently different for me. They're just manifestations of being human.

It's necessarily anthropocentric because I am a human but there's nothing exceptionalist about it. As humans, technologies are the things we create. And spirituality is a form of expression that is not represented in equations. Although even in equations there is a form of spirituality.

The world we live in loves compartmentalizing things. And technology today is also not just technology but also products. There's a whole system of capitalism and selling devices. There's a difference between a technology and a device like an iPhone. And that has a big impact on our spirituality.

Peter Bauman: You've described rituals as algorithms for the body and the mind: ritual as a protocol, running for hundreds of thousands of years.

Memo Akten: Obviously we designed or evolved rituals before we had the word algorithm. I had the realization that rituals are algorithms for the body and mind and that they allow us to reach a certain mental and physical state that would not be possible otherwise. That's literally what rituals do. That doesn't mean that rituals are only algorithms but that's one lens into what they are.

Whenever I teach a class on virtual reality or history of art and technology, I start with Werner Herzog's Cave of Forgotten Dreams, the Chauvet caves, because they were doing projection mapping. They were painting on the reliefs, they were using the geometry of the space, they were augmenting reality, they were painting animals by where streams would come, they were using firelight as a way to animate space. They were doing chronophotography (multiple exposures to convey movement) and they were telling stories. You go in there and you can see that they're not just dumb apes painting what they hunted. There's a story there. We don't know what that story is but it is just fascinating.
‍

‍

Peter Bauman: Superradiance, your lab with Katie Hofstadter, works on entangled intelligence and planetary consciousness. Do you have a philosophy of the mind that you build these works on, something like panpsychism? And could a machine ever be conscious?

Memo Akten: Superradiance is both an artwork and also the name of our research lab. It's a collaboration with my partner Katie Hofstadter that started in 2024.

Superradiance isn't actually panpsychist; it's more Gaia. The key difference is that panpsychism is making an ontological statement, that there is something that exists as a base layer of reality that is consciousness.

Gaia, at least the way I interpret it, is more a worldview, and that worldview is that the planet is a living system. Obviously this dates back millennia to all kinds of shamanist, animist, indigenous cultures. But in more modern times, James Lovelock and Lynn Margulis’s Gaia hypothesis is from the '70s. It's a worldview holding that we are not living beings on Earth but extensions of Earth.

But I will say that after many decades of being a computational functionalist, I have given up on computational functionalism. I’ve moved towards something I think I’ve always believed to some degree: that there is something that's missing in our most fundamental levels of reality which is linked to consciousness.

Self-consciousness, the model of the self as an agent in the world. I still believe that's largely computational. What is irreducible and not modelable by computation is what philosophers call qualia, the quality of any experience. If you isolate that I am now in that position where I don't think qualia is computational. I don't think it's software. I think it's a physical process. It's plausible that we could build artificial consciousness but of course we need to understand how nature is doing it.

The example I give is photosynthesis. When I say computation, I'm talking about the Turing sense, which means complete substrate independence. You can't program photosynthesis and create sugar. So it's plausible we can build computers that are conscious, the same way we could someday build artificial plants that photosynthesize, if we understand the biological mechanisms and then have the technology to build those systems.



------



Memo Akten is a multidisciplinary artist, researcher, and computer scientist working with emerging technologies both as creative medium, and as subject of critical inquiry. He is Assistant Professor at UC San Diego and co-director, with Katie Hofstadter, of Superradiance, investigating entangled intelligence and planetary consciousness.
‍
‍Peter Bauman (Monk Antony) is Le Random's editor in chief.