(audience chattering) - [Announcer] Welcome everyone to the Pen (uplifting music) (audience applauding) (uplifting music) (music stopping) - Welcome to the Penny Stamps Distinguished Speaker Series.

My name is Christina Hamilton, the Series Director.

Today we present AI Artist and Designer Refik Anadol and it's so good to see you all today.

You're gonna be so happy you made it out in the rai This is really a treat.

I truly believe Refik is t who's doing anything significant with AI in artwork.

So you're in for a treat.

Today's event is presented with support from the U of M Arts Initiative and Series Partners, Detroit Public P-B-S books and Michigan Radio 91.7 FM.

Just a couple announcements before we start.

This Friday, 6:00 PM that's tomorrow at the Stamps Gallery around the corner, Capture the Look.

Stamp students, Nina, Rochelle and Brian Wilson will be in conversation.

They're gonna present their work.

(audience cheering Yes.

They've got some frien but they are going, also goi and have a conversation exploring the intersection of fashion and photography.

6 to 8 tomorrow night, Stamps Do n't miss it.

And then remember we do have a wi th Stamp's Emeritus Professor Janie Paul at 3:00 PM that will be at the Art Museum in the Helmet Stern Auditorium.

Janie will be there with two formerl to discuss the significance of art in prison.

And then of course, we're back with Cannupa Hanska Luger.

Reminder, do silen and we will have a Q and A today.

There are the microphones at the en So do line up when we get to that point and you can ask your questions.

I know you're gonna have s Now for a little background on our guest.

World renowned media artist, Refik Anadol is a pioneer in the a and machine intelligence.

Coining the terms AI data painting, AI data sculpture and latent cinema.

He has been reflecting on new multisensory forms of narrating collective memory in physical and virtual spaces, encouraging us to rethink our engagement with the physical world and the creative potential of Exploring the relationship between the human mind, aesthetics, machine learning, and architecture.

He asks if machines can learn or process individual and collective memories, can they also dream or hallucinate about them?

Anadol's work has been exhibited at venues across the globe I think many of you in this audience today may hav his incredible piece that is currently up at MoMA.

It is also the first AI piece that MoMA has collected.

I believe it's called "Unsupervised."

He's been at the Centre Pompidou, Venice Architecture Biennale, Hammer Museum, Ars Electronica, and the Istanbul Modern to name a few.

He's received too many awards to name, including Lorenzo il Magnifico Lifetime Achievement Award, the German Design Award, U-C-L-A Art and Architecture Moss Columbia University's Breakthrough Storyte And on and on and on.

Please welcome to the stage Refik Anadol (audience applauding) - Hello.

(audience ap Thank you very much.

Good eve I'm so h and thank you very much for listenin I'm really happy to share our studio journey today.

I'm coming from Los Angeles, our city is in Los Angeles the journey that I have been hopefully share with you tonight can inspire, bring joy and hope.

I have been teaching at U-C-L-A last nine years and this year I took a sabbatical to make this kind of travels I'm like a bird traveling too much, carefully, to share art work.

So I am originally from, really coming from like a different background I guess that brings new dialogue to the technology, scienc So I'm originally from Istanbul, Turkey, where I born.

The city is very unique for many reasons.

It's literally geographically where the wes and east connects.

It's literally, emot contrast, feelings of left and right and future and past and it's just like a full of cultures.

And I think I learned many things about life there.

But I'm a truly nerd that I love computers.

I have been playing with machines for since I was eight years old.

I got my first computer maybe, you kn I dunno how many people play ga I'm sure it's a very common.

Okay.

So yeah, so I'm a I love machines.

I love to imag But I think what really transformed my mind is this movie, maybe you know, "Blade Runner."

It's one of the most I think, inspiring sci-fi movie.

But as a child when you watch something dystopic, I don't think we see dystopia, we see utopia still, we still see positivity, we still see creativity.

And when I look at this image, I remember clearly, by the way I didn't kn My cousin translated to me, like I hope he d (audience laughing) What I l feature of the buildings, where the buildings can or you know, drones and many others.

Literally inspired from this, in my undergrad I transformed in to like these canvases.

I start doing projection mapping, Bu t you know, 15 years ago it was not that common.

So the idea was if it's possible to turn light into your material and what happens if a building becomes a canvas and what happens we transform this canv into like some sort of like a dynamic forms and shapes.

And this became very funny, a lot of body work like ea and I learned a software called V-V-V-V, you can look at online, it's a free to create many things.

It's a visual programming language that means you d even though a J-G-P helps yo but you can still use without any sort of thinking too much but really connect the dots together.

And to me it was a very powerful moment of like thinking buildings as a canvas and as a child I always believed that.

What where we are living in, I mean this building is historical but what happens it remembers tonight and what happens it can somehow communicate with us.

And this was like always the idea and there's this incredible philosopher, he says, "Life can only be understood backwards; but it must be be lived forwards."

I believe that what really does And right now everything around us, the machines, softwares, hardwares, they are all communicating But to me, when we think about data, many people believe that they're just numbers.

But I see something like a memory.

A memory of a human, memory of a nature, memory of a thing such as AI or machines.

A memory that I believe that can take any shape, any form, any color, any texture.

So 2011 I start this journey with data and beyond.

If you think about humanity, we are all connected through systems, through software, through hardware.

This was one inspiring topic to me.

The other very inspiring topic to me and how artists trying to imagine light, shadow, like the edge of a building or like just materials.

Just things that artists always like imagining, going beyond what is physical or what is virtual.

At the moment, because ou I think we are literally in between.

We are sometimes finding ourselves thin about technology so much, but technology is transforming And I do believe many of you like this, that you don't know what is controlling whom and whe (audience laughing) So this to imagine what exactly Al so another amazing moment, how exactly we are in this new dimensions.

I'm a hundred percent sure, generalization of this room in many places like this.

But for me this is a place to imagine how can we go bey And I found that actually there's a much bigger questi What does it really mean to be a human in 21st century?

This beautiful creature learns how to use Instagram in just 10 Likes, not comments luckily, but does many things.

So this was a fundamental question that I thought that I have to di It really learns very quickly.

So, and to me this cannot be done one person.

It has to be a collective thinking.

It has to be not like an egocentric thing, it has to be something much profound.

So I moved to Los Angeles from Istanbul 2012 and got my second M-F-A degree from U-C-L-A Design Media Arts De And the dream was opening a studio, an art and design studio, uses technology, uses complex scientific discoveries, collaborate with scientists to make invisible visible.

Right now, we are 20 people can speak 15 language and 10 countries in one studio.

And we are Architects, AI Engineers, Data Scientists, Designers, Programmers, like really people love nature, love culture, love computers.

And last nine years, (uplifting music) I wanna quickly show you, we had worked so much about what you will see in this short clip, we profoundly look for how we can bring art, science and technology.

How can we bring the future of architecture and how can we bring art beyond just museum and galleries, and what happens if art is for anyone, any age, any background.

And what happe beyond the museum and galleries, hospitals, airports any other places like we were not expecting the art and I found it so powerfully connected with many people and last nine years our hope was public art.

Art for free, no beginning, no end.

Art that doesn't have a door or a ceiling or a floor, art that can truly transform invisible layers of data.

And I think what happens is, is the question is can we make inv Can we look at data as an new form of texture, a form, a sculpture, a building?

And you will see it in this short clip that we profoundly think about how we in an invisible way but still transform the facades of the building, transformed like surface o and just even use brain data, heartbeat data, wind data, polluted data, network data, data in all of its kind.

One thing we only focused, we never used personal data, we only fo And nine years was a long journey, when you think about lots of computers and canva millions of pixels.

And today I wanna show you a that I found that very inspiring, where we push the boundaries and how we can go beyond that and share the journey humbly with you, you will see robots that painting, printing, you will see human br you will see heartbeat of a human, like all that kind of in But one thing that really changed my practice, so 2008 when I was studying undergrad, I coined the term data painting.

It's the first time I saw that a sen and how it was just simply plotting this line and points.

And that was a day I remember create my first keynote about data painting and it's almost 15 years and I still don't feel that I know what I'm saying.

And it's still very exciting to say that seven years ago, 2016 Febr something awesome happened and that was a time this journey, kind of change, (music stopping) and this is what I'm calling it, Machine Hallucinations.

As we all know, AI right now is powerful in our life.

And how many people using large language models, J-G-P-T or something like that?

Like I mean right?

So, AI is with us.

So seven yea because I was so fortunate to work with one of the, let me just use this, it's probably much It's the most exciting day for and I was at literally San Francisco in a event called Deep Dream.

This image on the left side you are seeing, seven years ago just stopped the intern It was the first time people were seeing this weird image of hallucination that AI train on millions of animals and engineers were with or without knowing, find a way to create an image.

And this was the first time AI image became pu The group at Google thought about this wild moment and said, "This has to be also has an artist in thi not only engineers."

And I'm gr and my team become the first artist and res To me this was a great opportunity but I didn't want to just ruin t as a gamer, dreaming about one day machine can become a friend and now it was happening in front of me and I was deeply inspired by this incredible story.

If you're interested, the idea of that every single info There is no gates, there is no whatever's, just every information in life in one simple place.

Utopia, but the question was can we achieve this by using machine learning algorithms?

So 2016 we generate this immersive environment in a library, like literally open source library and physical place.

My question was when I go to a library, which I love so much, I don't really know like wha Even the best search like tools doesn't like show you or tell you in that specific domain what you need to look for.

And for this we took 1.7 million documents of library, create this environment.

And when you go inside this library, you just enter in this kind of a future of a libra where you can real time look at everything but not typing anything, just looking at it.

And we learned that nine years of work, of researchers were actually more interest when they compiled together that you can go and see everything and start looking at things that you never think about it.

And there was a new ways of an d I think that's one of the reason this project was the very firs and in public space and of course in VR, very eas Like what happens if you don't have a physical space?

Can we go to a library?

Can we look at millions Can we learn about it?

So we know The question was, if a machine can learn, can it dream?

If it can dream in a library where we learn what is real, who will define what is real?

So again, (music stopping) seven ye On the left side, this is the very in this archive.

These are very early generative by Ian Goodfellow, and he basically i one knows nothing, one knows ever and they try to communicate and creating and one tried to guess what is it and the other like in this like an examination.

And at the end you can look at this dialogue and pause it and work with it.

And this really to me inspiring to look at how machine learns and what happens if you On the right side, we took this inf If one day data becomes a pigment, I don't believe it has to dry.

I think data doesn't need the antonian physical quality of life.

I think data can move anywhere in any dimension, any direction.

It's just So this became a body of work that Last seven years, we are res Last seven years we download more than 4 billion images.

I believe it's the largest data every collected from a creator perspective, artist perspective.

And we train more than 400 AI models.

All look different information, nature, culture, space, time, architecture, things I hope belongs to everyone, not for one person.

And this became a massive research and last seven years you can see even those similar movements, they are actually very different becaus each of these paintings are and I figure out that this can be a painting, cool, but what else can we do with this pigment?

What else can we just take this pigment and the light as a material and go And if machines could like simulate unconscious, like conscious events, like what else we could work An d in the last seven years you will see different patterns, different colors and we are so grateful that we work with NASA G-P-L, Google, Intel, I-B-M, Siemens, but we didn't like do anything commercial.

It was truly just looking at algorithm, research and what else can be done with a complex information.

For example, what you are seein that is AI reading Rumi, Persian poet's poem, "Life."

And in one of data that is "Masnavi," and we let AI read this and in 19 languages.

What to me inspired me how AI sees the world and how it like understands.

And what we are seeing here on e emotion, one concept.

Similarities the distances completely machine made, aesthetically human intervention.

It's a true human machine collaboration.

The other part is Mozart, his entire life, 1784, 1791.

7 years of his body of work all l not Spotify, not other software.

What really inspired me is there's a world we don't see, we perceive and there's a world that we can use machines to reconstruct new ways of seeing.

Of course as Carl Sagan mentioned, "Imagination will often carry us t that never were but without it, we go nowhere."

So I took this concept that what happens if you think about a little bit more about the idea of embedding AI into architecture.

Now we have an AI that can learn and dream and what happens, we apply this model to around us and we did many projects, especially focusing on projection mapping, like this is a But what we did, we looked at a hundred movies, played in that space and then led AI to look at every single script of every single movie played in that movie theater and then we transformed th There's another like concept that I thi as an undergrad I was dreaming th with Frank Gehry, Zaha Hadid, Tadao Ando, Sana, Gaudi architect that I I do believe their buildings are not just buildings, they're like sculptures.

We go inside to li to just see the beauty of art.

And I'm happy to say that I was able to work with one of the most inspiring building in Walt Disney Concert Hall, home of LA Philharmonic, the building designed by Frank Gehry almost 20 years ago.

And the building is very important for LA Philharmonic and for hundred years LA Phil has bee and to me as a student, I was so inspired that one day, l like downtown Los Angeles, this building perhaps one So I challenged the institution and they challen they said for a hundred years, we wanna do something Not only fireworks, we wanna d an entire hundred years of every single information, every music ever played, every posters, every video, every image, everything like 77 terabytes of every music th And the question was if machines can learn or process memories, can building also dream, an d involuntarily remember or make connection between multiple people dreams.

So we took the entire building data from the Frank Gehry's origi and transformed the building into and the building was 20 years ago designed, so it took so much time to like 3-D model the surface again like reconstructive topologies an And when we think about projection mapping, it's truly looking at a lot of li kind of understanding how to look at the building as a surface that is not just a 2-D space.

Like we go 3-D spaces, we do many different worlds.

And this was the beginning point.

I wanna show you a short c to just show you what happens if a building can remember more than a hundred years of information.

We use 42 projectors, which each 4K, which is like kind of 13 IMAX movie on a building.

So let's maybe have a quickl (suspenseful music) - [Audience 1] It's okay.

(suspenseful music) (suspenseful music) (suspenseful music) (suspenseful music) (suspenseful music) (intense music) (intense music) (intense music) (intense music) (intense music) (music stopping) Master of Fine Art's Project.

Thank you.

(audience applauding) And, thank you And what was very important to remind you that even though it was an art in it was just before the pandemic and the institution was able to raise s and that institutional funding allowed the institution also go through the pandemic.

It was really powerful to see that art can truly trans when it's done right.

So there's a huge potentia I believe that one of also exciting project that changed our journey was for New York.

I love New York so much but I never though we can download 113 million images of and let machine to create almost realistic New York.

And most importantly the question was can we step inside these machine dreams?

We saw it in the building, we saw it on a painting, but can we literally exp in a three dimensional environments?

So five years ago we created this insta in New York.

(peaceful music) It's a 18 channel projection that you can that two dimensional worlds without wearing a VR, AR or XR.

The exhibition brought more than half million people together and it was one of our most well-known audiovisual performa that start a new movement.

So what we're seeing here is a 16 differ Each chapter is exploring different data sets of New York, the radio stations, TV stations, the building data, architecture, natu Explored New York from a complete different perspectives by two channel sounds and 18 channel projectors that are complete transforming in Chelsea market, a former you know, leftover place.

And that was really interesting to see that we can go beyond when w about architecture as a canvas and think about gallery as a whole canv And the (music stopping) other ar last year was in Barcelona, in He 's a really wonderful mind, I hope you can research his li He's a truly like pioneer of p truly inspired from spirituality, inspired from nature and combine them toget His craftsmanship, his legacy is extremely unique a and thanks to the building owners they invited to say what happens if you look at his li from the AI perspective?

Can we honor his life, his le And if we look at every single P-D-F ever created, we look every single driven he ever made.

And if we look his life in general and we received beautiful data sets, close a billion data points.

But what was to me inspiring is, can we talk about living Th e idea of him building receives the life such as weather data and impact itself or show the invisible patterns of information by its own facade.

To make it happen, we take the origin By the way the building is UNE meaning it has to be preserved pe and they use an amazing LiDAR scanning of the bu and that transform the building facade into 1.2 billion small molecules, parti The whole building facade was perfectly preserved.

And then we put a living weather station to get the data from the building and create a custom software that real time l to weather patterns of the building and transform into artwork.

So what we are seeing here on the left side, the humidity in the middle, rain, on the right side, weather patterns We were able to create kind of a living facade that transforms the building into a living architecture.

I do believe like coming "Blade Runner," inspiration, I think it'll be most likely exciting to think about architecture beyond the frozen form and see beyond.

Now I have one more short clip for you to show the question of what happens.

This idea transforms the real building, the facade of the building into a living architecture.

A short clip to share with you.

(eclectic music) (eclectic music) (eclectic music) (eclectic music) (music stopping) Thank you.

(audience applauding) I dunno if you can see the image, but there was 65,000 p It was the largest audience ever came together in Europe and I think people literally came to hug the building.

It was a really super emotional, very special night.

And I think it's very possible to say that we can connect the heritage the past now and future if it is well cu well taught and bring the heritage and the future together.

And I think there's a very much big room to explore this.

And lastly about this topic, maybe you know this building in So I'm so happy to say that I'm the first artist with U2 and then Aronofsky, and transforming this beautiful, jus I dunno how to explain this, but this aff 516 feet wide, 9 million pixels that we are seeing transforming into the machine hallucinations So this building is one of the most amazing forms that inside there's an amazing like performing arts center and in the exterior you can use as a like an artwork.

And I'm so happy to say that a we have the first and let the machine to dream, Hubble Telescope, I-S-S Telescop and we are also visualizing real time the invisible wind pattern of Vegas.

So I hope you can see it, it truly is sp And lastly, I think a to share today is the I think, at the moment, we clearly see that generative AI, a topic that is completely changing An d some people believe that AI will profoundly transform the what is reality, what is creativity, but I also believe that it'll enhance our thinking, the idea of enhancing the mind and the body.

I can't draw properly but I know how to compute geometry by code.

I know how to imagine colors and forms.

I do believe there are many people like me using computers to make creative things.

Of course when it comes to arts an the journey's kind of classic, you know, education systems, art schools, there are galleries, there are museums and so on.

So I'm not coming from a classical ar that or art education, I don't have a gallery represen I'm independent.

So that means that are in that your own journey.

And there's this funny joke when so if you are a dead artist that is not living, your best chance is living or being in the lower museum in Paris, if you're a living artist, the best chance is being And I'm happy to say that our show is at MoMA and I'm alive luckily (audience laughing) and I'm happy to say that we finally achiev one of the most inspiring experiments at MoMA (uplifting music) and I'm Paolo Antonelli, Jan Postma and my hero and te Casey Reas, who invented processing with Ben Fry.

And he completely transformed I think creative programming for entire industry.

For MoMA, the question was, all this archive project that you saw before and it was in the pandemic two years ago when the museum was closed, the museum challenged me and my team like like what happens if you take the entire MoMA archive, 138,000 artworks of beautiful minds discovered incredible pioneership but never tried to mimic them?

Never try to mimic a Monet or a Van Gogh but try to think about what happens if AI can become a research tool, a thinking brush (music stopping) that can think about It took us almost one bu t never try to mimic the artist but try to find the algorithm that generates not real, not surreal, somewhere in abstract space that can generate infinite artworks.

And this was a pure challenge and a lot of discussion with historians like thinking about how we can achieve thi that like mimics the art making, chance, control, but most importantly living architecture, right?

We discuss, can we make a living artwork?

Can we make it always alive and responding to us?

So I'm happy to say that this is like gun lobby in New York and it's a 28 feet by 28 feet media wall, super custom, literally 1.5 millimeter every pixel that can create this digital pigmentation.

And the thing that this project is unique is these devices, we have a camera and a microphone that real time can sense the audience l sense the audience movement, and plus we have the weather data.

We can let AI to listen, to see, to feel the weather data.

But what makes the artwork also special to me, we demystify machine decisions.

So if you go to MoMA, there's a one de that can tell you why exactl where the color comes from, how weather influences the movements and why in the morning maybe you know, students are in the museum very loud at late afternoon, very calm and so on.

So all this information demystified and I think one of the interesting part was this part, the other interesting part is we were able to use a supercomputer, one of these G-P-U based c and custom hardwares that on the left side AI is c and on the right side, on top of it, human intervention, not machine.

Left is machine, right is human.

Truly, and that connecting.

So when you see an AI art, people are thinking like it 's like a year long research, aesthetics, choices, movements all connec and the result can be very different.

For example, these ar from the last years of some AI dreams We don't capture so much because we let machi And I'm happy to say that the mus from all ages and backgrounds and at the end of the day we were able to create three artworks runni in this I guess, canvas.

The one on the left and the middle, completely real time, never the same and n And the one on the right side is a library of vis hundreds of them that are completely reconstructing itself based on the reality of life.

And here's a quick snippet of how it looks like.

(uplifting music) There are 300 million small molecules and each time they are tried to reach this data pigmentatio The music is also real time, AI made, like the LA Philharmonic Project in real time.

We have a custom software, looks at the movements of the painting that can trace the movement and transcend into this area, I guess Psychoacoustics soundscape and the project received significant audience and we really questioned this in such an incredible museum and with all this incredible wo this transcending situation.

And I'm working with an amazing Professor, Adam Gazzaley from U-C-S-F he's an incredible Neuroscientist.

And we asked the question lite So look at this pattern, why exactly this is happening?

And we learned that by using E-E-G devices towards 16 never see before, 16 saw before.

And we let people to like look at the work and measure the quantification of like the feeling, not just talking but really scientifically looking.

And we learned that there is a high chance that artwork is creating a flow state, a flow state that creates loss of programming language like people like using in like or in life in me There's a chance that maybe in a couple of years you ma that we may be creating artworks to ca and other, you know, life is complicated.

Perhaps we can find a way to calm down us scientifi by using art.

So that's And I'm happy to say that the artwork is rig in the permanent collection, meaning not only for me bu people thinking in digital arts and AI, I think future is bright.

And last topic that I wanna I do believe that humanity became completely differen Web 1, Web 2, Web 3, Quantum computation.

I think we missed the point that nature is t most incredible technology we have.

So last five years, we dedicate our research to nature, to preserve it, to understand it.

For example, this is an AI trained (uplifting m on 75 million flowers, 16,000 species that publicly available Smithsonian data.

This AI can create any type of flowers that exist in the online, in the archives of museums.

But it was exciting to create an AI that generate this, you know, forms that we can generate sound from nature.

But the question was could we smell this AI?

Is it possible to go beyond screens and the music?

And I'm happy to say that we found a way to do that.

So this is a machine we created.

So imagine AI knows half million scent molecules, that it can look at the real time image of a AI dreaming, that AI knows how to like create a smell and we let them create this that I feel (uplifting music) so fun to step i the machine dreaming these realities.

So what we are seeing is AI constantly dreaming these flowers and when you open the door you can feel different, 14 different molecules based on the color, the speed and the type of a species.

I found it so inspiring when we go this new world, still appreciate nature, not replace it, but still think about new art forms Th e piece is traveling, I think was in Dubai and now hopefully in US coming in mostly New York first and then hopefully you may have a chance to see this piece.

And also last month, sorry, two months ago, (music stopping) we finished anothe Like we just amazing look at all the like landscapes, beautiful like mountains and spaces.

This is AI can create dreams of national parks.

So we are now researching how can we think about national parks as a place to remember and for people who cannot travel perhaps but still appreciate the beauty of nature.

And this AI can really create probably any form of national park that has a representation in the archive and the music of the wind patterns and so on.

And lastly, I think this project was really inspiring.

As you know when the weather changes, when the nature starts to like disappear, there are certain places like glaciers or corals underwater first reacts.

Unfortunately corals are rapidly dying and GAIA invited by World Economic Forum, which is where the world leaders comes together every yea And they challenge us to say, (peaceful music) what happens if we use AI, not just create shiny beautiful pixels, but what happens if we use AI to preserve nature and help engineers and scientists to look at an image and perhaps reconstruct what we lost such as corals.

So we use 135 million coral images underwater, let AI to look at the patterns and create almost realistic corals.

And by doing that we were able to create realistic coral images to take attention of the World Leaders by Also bring the attention of the engineers and researchers who need support and we gave them a tool, not just the beautif and inspire and recreate corals.

And I'm happy to say that, what you in Miami underwater to reconstruct real corals to generate corals.

So these are not real corals, but details are so high that they can even create ecosystems when they were dying.

But there's a high chance that AI can be used to reconstruct what we lost in the nature.

And what is important is, in a place where politicians are co talking about separations and wars a we were still trying to take attention.

That is for everyone important.

And I'm happy to say that when it's done when it's presented right, the world listens and th that AI can be helpful.

(music stopping) And I'm happy to say last Febr pretty much the artwork get an entire world attention.

And right now more than five institutions are working with this AI.

So I'm happy to say that th e work of we do in different cities, different places, brought many people together.

I think we achieved somehow our idea to art for anyone, any age, in the background.

In our last sessions of our research that I guess in this keynote wanna share is So our next dream is, at the moment many AI researchers are looking for J-G-P-T, Stability AI, Mid-journey, all kind of these tools are amazing and respect but they are just product and services that are most likely creating value for certain people.

And we are trying to find what happens if use the same technology and go beyond that, make a gift for humanity that take cares everyone's attention and make something meaningful and purposeful.

So right now, we are researching large natu Imagine every single flora, fauna, fungi in the world that you can type and find maybe the best teacher of the nature, but the it is preserve the nature and make something powerful.

And I'm happy to say that an ethical data, it's ethical data from the institutions that we love, we respect and researchers who have been researching with that.

It'll be a gift to humanity and on top of it to much inspiring, hopefully important, we are physically traveling with the camera, with the microphone, with the drones, with the climate devices off to 16 different rainforests around the world.

And I was last of the most amazing places I hope you can trave It was in Amazonia, (rain falling) in the heart of Brazil.

The nearest hospital is 14 hours.

You have to travel through this incredible river, no private jet, no helicopters, nothing.

You are with the nature.

There is no way to go th your body, your mind, your soul, and most importantly hosted by these incredible people called Yawanawa.

They are a tribe living in the forest for thousands of years.

And they inspired, by the maybe you remember the trees and the mu They're one of those like inspirations for the mov This is Chief Nishewaka and his partner Phutani, on ly 17 people can speak their language but they live 120 years.

They don't have diabetics, they don't have co they don't have any problems that we have.

But what really inspired me is, if you think about d we have to understand where it comes from.

We cannot just think about nature and put images and sound and just think that it's ready I do believe that if we think abou as a living organism, which people I think it's also hearts of humanity.

I do believe that it's possible to connect with dat with the people who are living in it.

And by the way, if they don't exis those intelligent beings, biomes, may disappear.

And most importantly to me, to connect with them is im Ask them what else we can, they inspire me so much.

Like how can I inspire you as well?

These are young Yawanawa people and they're unfortunately don't have a don't have a museum, don't have So I asked them when I was living with them in like several weeks in the fo like deep in the jungle, like what, just what is the dreams you have?

And they said, "Of c We wanna have a school and we wanna have like a village, where we can con with all t And for this reason we use AI and ask young artists of the Yawanawa who couldn't go to school but they can draw.

They have amazing cultures, they have amazin We ask them, listen, just make a project together, co-creation and create its and completely give it to you and make your dreams happen.

So they give us 13 beautiful paintings and we use AI and these are their beautiful and we use their fundamental st the fire of inspiration.

And we use AI and let crea by blending artworks.

It's so exciting by the way, like Skype calls, FaceTi Zooms with the forests with, you know Starling like it's and so inspiring.

(uplifting music) And we also put a weat from the building and we collected data, real time from the forest, which is so unique and their be and generate 1000 artwork and create a philanthropy project.

And last month we raised $3.2 million for their village.

And this money without any government, any bank, any institution.

By using block And now they are building their, hopefully starting their museum, and their village to bring 3000 more tribes together in their village next year to celebrate.

So I think art can be very powerf and empower the people who may not have an access but we can create.

I wanna finish my talk with these beautiful words (mus of the amazing wisdom man that and he said "It's new times we are living now.

Time for forgiveness, time for love, time for spirituality.

It's time for human to To the Earth.

To our hea To learn, to To make alliances, join forces.

This is the moment!"

Thank you very much for listenin Happy to get your questions.

Thank you.

Thank you very much.

(audience applauding By the way, the Dataland, the Dataland is That's the name of the place.

Thank you very much.

(audience applauding And for yes.

- [Audience 2] Does this work, how do you compile How do you compile these massive data sets, you were throwing out like some insane num - [Refik] Yes.

So the, - [Audience 2] Okay.

- [Refik] So the with gia with cloud computing.

We have a two petaby but it's not that enough wi th giant data sets, right?

Like you know, mil So we have to use cloud computing in certain decisions and, but curating data sets takes months like ResNeXt or Logistic 16, like we just go through all that of different image cog to really understand.

- [Audience 2] Thank y - Thank you.

Ah.

- Hi.

I'm just an avera as when you're in the studio or like how do you balance the work life and travel?

- I think so as a studio by the way, we are super hardworking people.

Like we love to be together.

It's a horizontal space by the wa Everyone, their own leader Like there is so it's a very different thin But what is to me inspiring wh en I'm not teaching, I'm literally a lot of like, I just work a lot like 16 hours a day, I think, to be honest.

I'm super hands-on in ev and like hardware and software and design.

Like I'm just really love to be a part of it.

That's why it's a humble studio and a small studio that you can like in every single, you know everything about what's going on and understand why and how a So a little bit too much maybe, but I think it's where the mind and with the work that is very important I think.

- [Audience 3] Thank you, thank you.

- Thank Maybe here.

Yeah.

- Hello.

Your artwork very much speaks t and the way that you connect 3D, 2D and that like internet spaces just kind of all combined.

And you talked a lot about the economy of how buildings sort of speak to you and how architecture in itself is a visual space.

And I was wondering if this building itself, if you could think of one sentence that it would say to you, what would it be?

- Oh, that's a very beautiful question.

I think it's a really hard to find that, to be ho But I will say I do believe the architects that are right now practicing should be definitely focusing on how to think about this, you know, I-O-T, right?

These intelligent systems and sensors that can like think about like us together, like being here or the things that we see and performing arts here.

So first of all, we have and I think that is a huge research that I don't know anyone.

Like I have friend and Harvard were l but they're like so humble I think the experiment has to be scaled perfe and mostly in schools and, you know, education spaces and first get data and really understand what this building may wanna say.

And I think that's a beautiful question that requires a whole news, but amazing ques Thank you very much.

I have n - Thank you.

- Thank - Hi.

Okay, I have two One, what blockchain technolo And two, what do you eat?

- Oh yeah, so I drink coffee so much.

I love coffee.

I don't, but the blockchain So as we all know, the non findable t but there's ups and downs, problems and let's just take them out.

As a technology, the idea of using we taught any middle people, right?

The information goes something that computes in the infinite scales that you cannot close it, right?

It's so interesting.

And then you can t in what time and you know, for which reason.

So to me that was like a really starting because I can't draw physically.

I imagine in my mind and and brings to life.

To me it jus there should be somewhere else where the work can be before like trying to feel it in physical world.

So we write in on Aterian Blockchain mostly, custom smart contracts where we can interact with them like in the Yawanawa Project or like the Gaudi Project.

We have living works that can be const in the blockchain, also in the cloud.

So we basically connect the bl and the cloud computing together so that like they interact with each other The work is not necessarily one piece of but it constantly changed an active like response to life.

So it's really interesting to think - Wow.

Well thank you.

- Thank - Hi Refik, I have - Yes.

- First, what are in order to determine whether AI can hallucinate in the first place?

How do you think that' in terms of measuring qualitative concepts?

- It's a great question.

I think th I don't use personal d just the high ethics borders that I That means that I only look for things I And that really keeps like things very, I think under control when it comes to And then from there, of course as an artist, I never use other artists like informa I always like try to create our own custom model from scratch that I know exactly what type of information in it.

And I think for a people u and division models, I think it shoul for the ownership of the idea.

I think it's a very important part.

This takes time and resource, - Right.

- And th I will d like the originality of the data comes from the artist de And it's very doable to create your data, train your model, fine tune your model and still enjoy that creativity from the mind, to type, or whatever, E-E-G or like telepathy, but you can some to like make that happen.

It's very important.

- Okay.

- Thank you.

- Also, I know that from U-C in creative programming, which I also did last year w - Amazing, the object base.

- So good, so So what do you think What are to get to like- - Game engin - Game engines are - Okay.

- Unity, I know V-V-V but but I think Unreal Engine Unity their both amazing And I personally love Unreal because of many reasons, but there's a huge potential to take the creative cod and the game engine world- - Right.

- And cr And I thin you know, self learning, learn to learn.

- Thank you for your time.

- Thank you.

- - Hello.

Hi, thank you for being here.

- Hi, thank you.

- You introduc on in one of your pieces and I'm curious, I saw that used in many of your works and I'm curious as to what the significance is of that, 'cause i - There's a lot.

- Yes.

So It's the easiest.

So I think wha fluid dynamics, I mean again, 12 years ago I started using Softim a software invented by Autodesk X-S-I, which was early studies.

And I didn't like hundreds of millions of part and they're responding cert What is to me is very inspiring is the i of like really fine tune these moments, crafting them by the viscosity, the length of the particle.

Like to me they're intel I mean something just like, okay, but it's actually very com Some softwares takes like in Houdini, 40 terabytes per just four minutes of simulation.

So it's a really intense computation like in the giant movies that people are using I guess.

But we try to use in like art making, even though they look similar.

But if you look at, you and different data sets, there's always one and I love that change over time.

And now it's also real time with the computation, the GP computation that has, you know, ups and downs and limitations or cloud computing or like, you know, Gaudi's building.

But there's a lot of room I in this aesthetics and I allowed this frame i it was a rejection 2011.

So I was thinking about put the give and they said like, it's super exp that hurt me so much.

I guess since around the work.

And I love the f because it really gives this idea of l the early example I show from the, like this painting that a human was almost going out, like that kind of a very like, you know, ancient Greek like hematologies I love that world so much.

But as you see, like, you know over time things just change if you have a, y deep, you know, research and look for.

- Mm-hmm.

Thank you.

- Thank you.

- Hello.

- Hi.

- I'm de I work for LinkedIn and I'm using and the tools we're building.

- Amazing.

- And we and we're creating things that have to be very deterministic.

And I was curious how you define versus dreaming.

- Yes.

- It see it's acceptable to because you don't have to - Yeah.

(laughs) - So tha And my other que like showing the black box and how AI works.

- Yes.

- 'Cause that in a 3D s - Yes.

- But AI - Yes.

- Not ne - Yes.

- And I to understan - Mm-hmm.

- Of how because every co "Oh, we can't explain how he did that because it's a black box."

- Mm-hmm.

I mean, it's a I wanna show one, oh by the way, this was the frame rejected 2011 that I co Yeah.

So.

I think this is like a, I think good examp So, what you ask is latent space.

Latent space is a place where machine learni And unfortunately the systems we use or the software and products and whatevers, they're trying to hide it.

That we don't see exactly how machine And I think this all happens in here, by the way, for everyone, AI algorithm not in three or four dimensions.

Like this is 10th time four dime of a machine learning algorithm.

So what you see here and a kind that we have to write in Python that on the left side, I'm usin that I can travel in this like imaginary t of 10th time four dimension and imagining like this is a th at you can bend inside the latent space and every pixel or latent points are kind of metals and this is some kind of like a magnet that it can touch them and An d so that's how I find the way to interact with latent space.

And this can be applied to any of course s It's very heavy research, not necessarily easy to use, but I hope that every single product and service has a way to show us where exactly that output is coming from.

That's one of the experiments we do.

- Thank you.

- Thank - Hi Refik.

- I'm a big fa - I just - Thank you.

- To Ann Arbor.

The weat I'm also a big fan of Blad so I found it like find it really inter But I just wanted to ask you, my question is, with your pieces of work, is there like a that you're trying to like incite from the viewers?

Like how are you sort of approaching that?

- Amazing.

I mean, to the emotion when you use computers, right?

Because sometimes machines a to create deterministic, like predictable world But I do believe that when we try to get out of our comfort zo that like.

There's this, I think that is can I think only be achieved when we truly a little bit get out of like this super deep comput For example, in my Google residency, they ask like, "Hey, like tomorrow you are meeting with the best engineer.

What you wanna ask?

Like, are you exactly-" Like I said, "Yeah, I' but I mean most likely it can into a boring mathematical talk What will be so cool if I can be an artist, AI engineer, neuroscientist, and a shaman.

So when we four of us sit together for a day, then it gets very inspiring because we'd start multi-dimens multicultural and go beyond the bias of the a and science and technology.

So these are the practice I'm trying my be to get out of the zone and really think from my heart, not just mind, because about the systems, the signs, and that's I think a practice.

And if you can, I think do that, there's this moment of touching the and the soul moment.

And I think if the a I think it's a truly success moment in my humble view that is So that's one experiment.

- So cool.

Thank you for sharing.

- Thank - Hi Refik.

- Hi.

- Thank - Thank you.

- I'm also a b as an ar in data and its translation through visualization.

I, you kind of, I, I've been really inspired by using like progra like Touch Designer- - Yeah.

I was just kind of curio how do you choose which components to use as like parameters within stuff like that?

Or just kinda like what, how do you kind of make the de like how much is AI incorporated into that versus how much human intervention is How are those decisions being made?

I'm just kinda curious - Yes, happy to sh I mean, amazing and V Four like or Maximus Com they're all like in So for me, the data like the Some data is about like, you the sensors, right?

Informat E-E-G data, like completely different la Like here for example, I'm like plotting the, this is the third two-channel E-E-G device, which can real time record the memory, the moment of remembering.

And we have another AI which knows what is the moment of rememb and it can look at a positive or negative memory, ri And once for example, like the let's say six distinct part of the br that where based on neuroscience, they say look at these if you wanna make somethin I really took them a and generate them into a Like really try to use science as a b but of course after a certain mile try to you know, bias box of like, you know, what humans find, but what happens like if machines can little interpretation, it's like a dialogue to be honest.

Sometimes super like explicit, But the spectrum is really based on the cameras, audience and many others.

- Thank you so much.

- Thank you.

- Hi, my name is Brooke I am really in love with the ty of work I saw Datavers and I thought that it was absolutely amazing.

- Thank you.

- And I if you mention that inter - Yes, yes.

Always.

Always.

Yes.

- If so how can we apply to - My email on the last like, let me go.

I mean happy to.

So every summer we are hosting many, many people imagining, wait, where is the, ah, here.

Let me write my email just on, Oh.

Sorry for the, this is the, this email upper right.

- Okay, cool.

- Yes.

So happy in this weird world we are in and if it inspires.

- It very much does.

Thank you.

- Thank - Hi, thank you so much for your presentat - Thank you.

- I was really fascinated by th at you implement photography and you mentioned that you accumulated millions and millions of photographs for your project And I was just wondering if these are stock photos and how does copyright come into play with these vis - That's a great question.

So what we never show what is real.

You will never see what's the sou We only show what that millions of images and all of there's one pixels connected together makes something.

So for us ethically, these are collective memories wi thout any password.

They are online.

That is and find it without any barrier.

Unless someone is hiding or private, of course they're different.

But if it's a pu it's a public data in the context of machine.

And we never show, never ever, show what is rea And that's a very beautiful experiment.

Last seven years and I talked with, yo the largest IP, you know, love years in the world.

Like they're just so excited to see that when mission drea have to go back, what was real, almost impossible.

So that brings this whole opportunities and same as like J-G-P-T or Mid-Journey, like it's a similar situation.

While those systems are much, much using real, real things that creates problem But in our case, we intention So when you don't use the algorithms as they're designed for, yo u have this hallucinations which are completely surreal, like going to a museum and inspire from a work, go home and paint.

I believe, it's a similar situation when it comes to like using self recorded data, self-trained algorithm, curated data and never sho and keep that cumulative feeling that is impossible to go back.

- Thank you.

- Thank you.

- Hey Refik, how are you?

- Good to see you.

- Good, So grateful for your presence.

- Thank you.

- Thank and make it a memora - Thank you.

- So I'v - Ah!

- And th - Amazing.

- Just w Also, I don't have a specific qu just wanted to see if you could share more information on DataLand and- - Yes.

- Also we have s to see if there's any alphas.

- No way.

So you from, okay, this is so e We have a 15,000 people in Discord.

So the web three community is but we don't like I just met now you, but I mean we are connected through like this, the systems.

And so Dataland, is it in recording?

I don't, I mean it's a museum.

We are imagining, it's a p and other artists who, you know, outliers, misfits or whatever we call it that couldn't like fit in this predictable But people who have dreams about data and AI or digital art in general and make incredible worlds.

It's opening in Los Angeles, hopefully- - Ah, okay.

- End of next year.

- Okay - It'll be the It'll be on Web 3 of course, there will be like a virtual ph ysical version, ac So it's very different than other proj and it's a lot of work.

So it started like a las we work with Google friends and any media friend Like we are trying to merge thos that they don't work together but bring them together for create one of the most never seen before experience in l - Gotcha.

- Very high level.

- Wow.

- But it It'll be whatever w But it'l to bring everyone together.

- Beautiful, beaut - Lovely to meet you i - Pleasure.

- Hope to chec - Pleasu - Hello.

- Hi.

- Thank I was able to see your work It was amazing.

- Thank you very - I was with watching all of the unique things that were being developed in front of them.

And from your talk I wanted to as about your AI data sculptures, particularly with Rumi's poetry.

- Yes.

- How ar upon the emotions of a poetic form?

- So it's a great question.

I mean when it c these are like, this So there are like a multiple AI models at t that are open source and I don' we don't use anyth we only look for open source op en source culture.

And they're like t I think Yumnet was for the Rumi and there are like multiple, I don't honestly rem but there are many open source, knowledge based models, not necessarily used for J-G-B-T or other things, but where you can classify the emotional cont of sentences, individual words.

So you can have a chance like to look at AI, let you let AI to look for vert or a semantic of like a sentences.

Here, this is all the words clusters which is 19 different language and there is 19 unique clust Here, we connect the sentences to look of like the meaning of a sentence which creates a different visuals.

Here we create both together and look both the word and the sent to generate some, like it's, these are li but I don't think there's a one way of doing it.

There's multiple ways.

Like in Mozart we were We use his repertoire and create like a chunks of seconds, milliseconds and let machine to listen.

Short ones, long ones I think in middle, but of course the short ones create bigger data sets but doesn't create the meaning because you know Mozart like doesn't dream milliseconds, But we try to let go much longer repertoire which make this middle, middle center experience.

So there's so much way of looking data from a different perspective.

- Thank you so much.

- Thank - Hi.

I was ju or predictions about how AI technology's like going to develop in the future?

- I think there's a high chance that it won't create negativity as people are right now thinking, I think negativity coming from a little bit hysteria, a little bit like you know, not lear Like more like just imagining without learning, right?

There's like this guessing without learning moment.

I think there's a majority of the pe that when I see comments, like people don't know how m that you have to put before mission I mean for me it's a 50/ 50% collaboration, nothing less, nothing more.

And that gives me the it 's a pure collaboration, right?

For us when we use those, you know, large language models, I don't know how many people really react that yet, but there's of course a chance of privacy and field problems.

If a machine really li what we eat, what we watch, what we read, what we buy.

I mean there's of course a challenge of like creating this bubble of re These are very predictable but they should be like in everyone's educational like background that knowing that the system that you are using has an AI behind it.

And I think when it's more exposed and demystified, the chance of fears will eliminate the hostel rise.

So that's what in our exhibitions, we always dedicate a screen wh which algorithm we use for which reason and that's one way of doing it.

- Thank you.

- Thank you.

- Merhaba.

- Merhaba.

- Thanks for coming to Ann Arbor.

- Thank you.

- We are Well I've been to two exhibits of First one was in Istanbul at Las Namasis and the second one was in MoMA, New York.

So it both were amazing.

- Thank you very much.

- Mesmer - I don't know how to explain, so you mentioned that the - Yes.

- Or are - Could - Yes.

So I mean we r just think 44 minutes average user, you know, is a very long time.

Like it's not a normal, it's not a movie, it's not like a performing arts.

So it's not, it is creating so of a positive pattern.

Again, thousands of mess when you combine this data, I think the artwork unlocked something, a feature that was, you know, So we try to like find the pattern what exactly happening.

So we put friends in U-C-L-A and do F-M-R-I machine, D-T-I scannings and looking like what happens if you watch a machine made works versus human What happens when you look at a nature that's real?

What is AI driven like?

We just correlate a lot So next year, there's a P-B-C documentation coming and it'll be like showing more and more behind the scenes.

But it's ongoing research.

I will say we look for t blood pressure, blood temperature data and M-I-T professor, amazing machine call that is used by the Rosa Picard of M-I-T, an amazing professor invented effective computing So like we worked with her at device an we work with like Adam Gazzaley, like we try to combi that we can quantify the feelings.

It will take time, but it's a happy t to record data so it'll be hopefully more historic for other art forms.

Thank you.

- Thank you for staying for so many questions.

- Oh please, please.

- And I wa so thank you.

- Oh.

- As well yeah.

I, you already mentioned the of computing power that's need to make this sort of work.

- Yeah.

- And I to the envir - Yes.

So if yo yes it is a huge clusters of But if you're fine tuning an AI, it's just 110 volts, you know, in an office computer.

There's like the scale is Generally in our studio we are using like classical like there's nothing extraordinary.

There's a really normal, I guess art and design And even at MoMA the machine we use D-G-X station, it's the supercomputer, yes.

But on like a two-two inte like it's not ne ridiculous thing.

The screen w this projector most likely the sa like a, you know, like a cinema theater level.

So we capture that as much as we can.

But when it comes to big data sets and when it comes to this like gia I don't know, 70 billion parameters or Yeah, they use a lot of data and a lot of like power.

The reason we prefer to use Google Cloud is exactly, carbon neutral data energy they're using.

So when you look at other like cloud computer soluti we couldn't find honestly, demystifying how much energy is using.

But when we started using Google Cloud, you can go online and check, it's like they h of the webpage and deep mind spend years to fine tune their power usage and they were able to completely 2030 carbon free and now already, already almost, almost there, leve So there's a way to choose the in stead of going to the more dangerous ones.

- Awesome, thank you.

- Thank you.

- Hi.

- Hi.

- Thank you so much for - Thank you.

- And for your work.

I've been inspired f by the Archive Dreaming - So good to meet you.

- Good t I'm curious about when you into the physical space.

What kind of drives the decision between like using infi versus like the framing technique?

- I think one is a data painting, one is a data sculpture.

When I say a sculpture, the room becomes a canvas and there explore like th because to me the library and informatio I don't believe just putting a floor just represents that feeling of like the big data.

So to me mirror represents these infinite p and potentials.

When it comes to paintings, I mo re accessible scale, like 85 inch under whatever inch like that.

Those like a TV like a really canvas, digital canvas.

The buildings are a whole different world, but when it comes to to me they are data sculptures.

- Thank you.

- Great to meet you.

Hi.

- You wo so I was wondering if you've e or if you already have like worked with like holographic projections on like transparent screens or even like reactive images on the projection that like, or interactive images.

- Yeah, maybe you saw that we've vis That's where we were like interacting with every single TED talk e That was one of them.

We did m To me there is still no holographic, you know, substance in the air really flies and that, you know, that didn't happen.

There's only like Pepper Ghost effect taking a 45 degree like light refraction that you put like a source here and then you see some, but I think it's exciting in certain conditions.

But it's not, to be honest, super innovative.

These vaticular screens are exciting that, you know, mimics three dimensional thing.

But I'm really waiting, truly a su that can compute, there where the hologram is.

No idea who will invent it.

But that's, I think the answer Good to meet you.

- Thank you.

- Thank you so much.

So have (audience applauding) Bye-bye.

Thanks so much.

(audience applauding) (audience chattering