How This Artist Uses A.I. & Data to Teach Us About the World

Artist Refik Anadol doesn't work with paintbrushes or clay.

Instead, he uses large collections of data and machine learning algorithms to create mesmerizing and dynamic installations.

[otherworldly music]- [Narrator] It's easy to get lostin Refik Anadol's surreal installations.But dig a little deeper,and you realize his workis made up of millions of tiny pieces,and that every single little pointrepresents a piece of datathat he's fed through a neural networkto show us his vision of our future.- I was always trying to speculatethis idea of, "Can data become a pigment?"At the end, the data is truly numbers.It has no kind of inner skin or skeleton.But what I'm trying to do asan artist is find algorithmthat can narrate the moment of data,kind of make thatinvisible moment visible.- [Narrator] Enormous data setsinspire most of Anadol's artand he uses machineintelligence and algorithmsto create visualizations,what he calls data sculptureslike this piece, called"Machine Hallucinations."He started by finding 113 millionimages of New York online.- [Refik] I let the machine learnfrom this entire corpus of data,and then we also find a wayto erase the humans from this dataand only focus on the buildings,nature, and environmentsthat is a collective memory of New York.- [Narrator] Once all of theimages of people were removed,Anadol was left with 10million pictures of New York.He fed them through amachine learning algorithmthat generates visualassociations as it learns.For example, when it sees multiple photosof the Statue of Liberty,taken from slightly different angles,the algorithm interpolates informationto help it create a moving, shifting imagethat represents the entirelife cycle of the structure.- Machine looks at thisinformation like a human being,but it's kind of morelike collective memoriesthan personal memories,because a building in New Yorkcan be explored bythousands of perspectives,from multiple angles,from a different time of the year.It's more like an honestmemory for a machine'cause it feels more totalitarian,and feels everything and everyonethan just one person.- [Narrator] Anadol coveredthe walls and the floorof a boiler room in abuilding in Lower Manhattanwith this machine-interpreteddynamic landscape.If you ask him,he'll tell you that themachine is dreaming.- When a machine learnsfrom outputs and memories like this,it can create an alternative reality.It look at the patterns ofthe trees, the buildings,the nature, the people,every single thing hiddeninside these image corpus.Seeing a machine giving a context of dataand giving an hallucinative outputwas something really inspiring.- [Narrator] Anadol created this pieceto celebrate the centennialof the LA Philharmonicusing half a million images,thousands of audio recordings,and hundreds of videos fromthe orchestra's archives.He fed all that intoa series of algorithmsthat turned it into theseextraordinary projected images.- [Refik] By using machinelearning algorithms,the entire archive of LA Philharmonicbecomes a shape ofthree-dimensional outputs.We were able to see all these data pointsbecoming together or disappearing,and making kind of new sculptures.- [Narrator] And the building itself,Frank Gehry's Walt Disney Concert Hall,became part of the sculpture too.- I was always lookingfor inspiring buildings,and Frank Gehry became my hero.And I rent a car, went todowntown, it was 2:00 a.m.The building's light was off,like there was nothing around it.And that night, I wasreally struck by an idea,like, "What will happen ifthis building can remember?"Like, "Can it dream?"- [Narrator] So he projectedthe machine interpreted archivesonto a smaller model of theWalt Disney Concert Hall.Once he settled on something he liked,he projected it onto thebigger, real life building.- [Refik] What we are doingnot a typical painting,not a typical sculpture,but more like an experiencefrom a near future.And when they come togetherwith sound, data, machine intelligence,and light and architecture,they have a new meaning,a new kind of symbiotic connections.I think I was really hooked upby the idea of, "A buildingcan become an interface."It can be a boring advertising,but it can be a beautiful interface.The ideation of a human instinct,becoming part of a machine,is really striking me emotionally.- [Narrator] Anadol and his teamused 42 large scale projectorswith an astonishing 50K video resolutionto create a dramaticdisplay in Downtown LAevery night for 10 days,a nod to one of his favorite movies.[whimsical music]- I was eight years oldwhen I watched the movie "Blade Runner."I was extremely inspired by the movie.In Downtown Los Angeles, abuilding suddenly becomes alive,and it has a cognitivecapacity of remembering,and it has a capacity of dreaming.So this was all very high-levelscience fiction narratives.- [Narrator] In thefuturistic dystopian films,exploration of more complex themesecho Anadol's interest in the relationshipbetween humans and technology.- You think I'm a replicant, don't you?- When Deckard and Rachaelcontemplates each otherand has a dialogue,where Deckard definesRachael as a replicant,it was a very interesting moment.Where a human defines another machine,that is kind of this,defining what is real or what is not.That kind of human consciousnessand machine consciousnesswere having its dialogue moment,that was truly inspiring.- You remember the spiderthat lived in a bush outside your window?- The egg hatched.- Those aren't your memories.Those are somebody else's.- I think, in humanity,we have very certain findingsthat transform who we are.When we found the fire,we cook with it, we create communities.With the same technology,we kill each other and destroy.And clearly, AI is one ofthe discovery in humanitythat has a potential to make communitiesor destroy each other.- [Narrator] Anadol finds datafor his machine intelligencecollaborations everywhere,and when he learned that when data iscontinuously being collectedat the Boston Airport,he knew he'd found a gold mine.He turned it all into this projectcalled "Winds of Boston."- I was always inspired bynature, as an inspiration,from the wind itself, the water,the nature, like a core nature,like aspect of motion,theory, and life in general,and I thought this can bean incredible opportunityto use wind as a data and as a pigment.So we first located our data sourcefrom the airport, Logan Airport.The wind is extremely importantfor the flight networks.So we took a one-year-longwind data of Boston,and this data consists of thegust, speed, and direction,and the temperature of the weather.- [Narrator] Anadol fed all this wind datainto a series of algorithms,then he built custom13-foot-tall LED screensto display the visualized data.- And I thought that algorithmscan be an incredible wayof visualizing thisinvisible pattern of things,and transform them intoa poetic-like motion.Because it's very inspiringthat a machine can findsomething interesting,that I didn't even think about it.And I found that the true collaborationstarts, happen there.And sometimes machine gives opportunitiesthat I didn't think about it,and I do believe that is really somethingthat is inspiring for humanity,that will most likely bringsome new kind of imagination methodsthat is not discovered yet.- [Narrator] The inspirationfor this project,called "Melting Memories,"came from a personal place.- I went to Istanbul, my home town,and my uncle just couldn't rememberwhere I am coming from.And apparently it was hisfirst stage of Alzheimer's.When this disease happens,we literally melt our memories.Our brain tissue disappears.But I also was very curious about,scientifically, honestly,what memory means.Like where do they come from?What is the cognitiverepresentation of a memory/- [Narrator] And the imageson these 20-foot-tall LED screensrepresent the real databehind that process.- And these moments arereally trying to give a senseof a tangible feeling of a memory.I know that we are not there,as technology that trulytranscribes memory,but it at least gives a glimpseof an abstract language of a memory.That's one of thosefeelings that is hiddenin those algorithmic exploration of data.- [Narrator] Anadolpartnered with scientistsat a neuroscience lab.They asked subjects toconcentrate on childhood memories,and recorded all theirbrain pulses using an EEG.- [Refik] The data itselfis literally pulsingin four milliseconds ofhuman brain activity,in the two location of brain,from hippocampus to frontal left lobe,and the algorithm symbolizesthe moment of remembering,and it is how poetic,moment of this firing locationin two locations of the brain.- [Narrator] Anadol and histeam created custom softwarethat transformed all of that brain datainto an artistic interpretationof the neurons firing.Another example of howwe use this technologyto describe life.- I never stop thinking about dataas a mean of like material.Sometimes it can be wind data.It can be wifi, be a little signal.It can be machine decisions.It can be CPU data, GPU data.It's honestly anything that machine needsto understand life,can become a materialfor this imagination.One problem we have tosolve today [laughs]is the visual discontinuity problem.- [Narrator] Anadol has built a teamwith expertise in different fieldsto help him fulfill his vision.AI experts, computer scientists,architects, and designers.- [Refik] Produce them, wewant to make them continuous.- [Narrator] And they'reworking at a new work flowfor a data-driven public artproject set for Portland.For this one, Anadol'sletting hundreds of thousandsof images of Portlandinform both the projectionsand the structure they'll be projected on.The team is using a 3D printerto build the structureone panel at a time.It will ultimately be a21-foot-tall sculpture.- [Refik] I found that we were stuckin this just virtual world.I found that we are in the screens,2D flat world of imagination.I was really looking forward,how we can take this machine consciousnessout of the screen andbring it to the 3D world.- [Narrator] Whatever's in storefor the future of Anadol's work,data machine learningwill be the foundation.- It's very clear that machinescan really capture our data,machines can capture ourdecisions, our memories,but it's not clear whatwill happen to them.And I think, also, clearly,the data we leave behindis the memories of humanity.My obsession with this relationshipis what else we can do with that.