Digital Arts@Google: R. Luke DuBois and Scott Draves

Uploaded by AtGoogleTalks on 10.08.2010

[Indistinct pre-chatter]
Josh: Good evening! Good evening, ladies and gentlemen, welcome to Google. This is the
second in our series of artist lectures at Digital Arts Google. Thank you very much for
coming. This evening we're very pleased to welcome Scott Draves, and Luke DuBois, both
of whom have pieces exhibited out in the lobby.
I hope you got a chance to see them before the talk, if you didn't, then in any case,
they'll be there after the talk as well. Digital Art at Google is a new venture for
Google. Google's general mission is to organize all the world's information and make it universally
accessible. Art is one way to organize information, and therefore we feel that computational art
is a tool that Google should understand, and eventually, should be using and helping people
to use, to present information to the world.
I'm not going to talk anymore, both because you didn't come here to hear me, and because
I have a terrible cold.
Scott Draves has been doing computational art for about 20 years. He's been displayed
at festivals and museums around the world. A lot of his work are examples of massively
distributed computational art, including the piece out in the lobby right here, and I won't
take up any of your time.
SCOTT: Thanks, Josh.
So I'm just gonna switch over to my laptop for a little presentation here, and what I'm
gonna do is tell you not just about this piece of art, but about a couple of pieces, a couple
different things that I've done over the past 20 years, and to sort of give you an idea
of where this came from, and maybe hint at where it's going. So, it all started back
in the seventies, when I first encountered computers as a, as a child, and it was, it
was love at first sight, I was fascinated, and started programming, started teaching
myself the programming, and started doing computer graphics, started drawing pictures
by writing programs. And that developed through college. I went to Brown University and was
in the so-called "Graphics Group" with Andy Van Dam, and that was a really rich period
in history, Scott Snibbe, who's in the audience somewhere? Where'd he go? There he is, in
the corner, was there, Brian Knep was there, two other artists doing, programming computers
to do interesting art these days. That was also a time when I really got into open-source
software. In particular I learned the emacs editor, written by Richard Stallman himself,
and really got enamored of this philosophy of sharing code and collaboratively in a distributed
fashion, sort of improving intellectual products. And so then I went to Carnegie Mellon University
to study Computer Science with people like Andy Witkin and Peter Lee, and I was studying
programming languages, and something called Meta-Programming, which are programs that
write programs, or things like compilers, and semantics, which are sort of the meanings
of languages, or the meaning of statements, and artificial intelligence. And it was during
those years, from '90 to '97 I had a lot of, a lot of rope. My advisors were very, were
very generous and supportive of my, of my research and also my, my artistic pursuits,
and that was really when the ball got rolling. And in, in particular, there was -- I had
an internship in Tokyo in 1991. That was where I had access to a silicon graphics supercomputer
all, all to myself, for the whole summer, and so for a long time I had been working
on these graphical algorithms, and I was always constrained or limited by what I could do
with my, my regular PC that I had at home, and suddenly with this supercomputer I was,
I was sort of freed from the constraints of time. It was sort of like a window into the
future and I was able to really try to -- I was free to find the, the real beauty inside
these equations, and inside the math.
So, the first thing - oops. I was talking too loud. (Pause) So the first project that
sort of came out of this process that I'm gonna show you is called "Fuse", and I wrote
it out, in, in Japan, on that trip, and it's quite different from work I have I have here
in this show, because it's based on input, input images, so it's not abstract. You actually
give it a photograph or, of a painting or of a, of a scene, and what it does is, constructs
a mosaic out of that image, and it's a mosaic where the tiles overlap. And what the algorithm
does is, is it does a search, and it searches for the best tiles that, that sort of fit
together on the overlapped portion. So it's a little bit like a puzzle, where the computer
solves the puzzle, and it's possible to have a target image. And so, an example of what
this algorithm produced is here -- unfortunately these projectors are really dark, I don't
know, you can barely see it on them. Can you, can you turn the lights down? Does that help?
You can also, maybe you can see a little better on the LCD screens. Behind you. Well maybe
if I just turn, does turning the brightness up on this screen make any difference? I don't
think so. But basically this is, I took a photograph of my face, and of, of my sort
of friend and coworker's face, and I fed those two images into this algorithm, and it produced
this as the output.
And you can see fragments of faces, but it's all sort of scrambled, so locally it makes
sense, locally it looks like a flesh, but globally [coughs], it doesn't make any sense.
So here's another example, where I took several pictures of my hand, in various configurations
and fed it into this algorithm, and it synthesized this sort of "fleshopod" landscape, and those
are -- It's also possible, like I said, to give this algorithm a target image. And so
in this case, I gave it the same picture of, of my head, my portrait of me as the target,
and I gave it just my [coughs] excuse me, just my eye as the, as the input image. And
then so it, it made a picture of, of my whole head out of just my eye. So this is the result.
It's kind of creepy.
And so again, these are images from the, from the early early nineties and sorry, I'm really
dying, can - Josh, can you get me a glass or bottle of water? I should have had said
that –
And so this - [sighs] was algorithm was published as free software, open-source, on the web,
a few years after I wrote it, after the web started to, after I made a webpage. And then
it, it was put as a plugin into the GIMP software, which is the open-source Photoshop clone,
and so it shipped onto millions of desktops all over the world. And the technique actually
was then picked up by the sort of traditional computer graphics research community.
Thank you. And so this is now known as patch-based texture synthesis, and there are hundreds
of research papers and people working on it, improving this and doing different things
with it. And you can go to to see a whole gallery of images.
One thing that I actually did a lot of when I first made this, was I used internet pornography
as the input images, so a little bit like Mark Napier's Pamela Anderson. I took images
of Internet women and sort of scrambled them, and created abstract art out of them. I'm
not gonna, I'm not gonna show those here, but you can, you can find them on the, on
the website. So, that was, so here's another project I
did, so that was "Fuse". This is "Flame". "Flame" is an open-source visual language,
and it-- I put the date on it as sort of 1991, but it's pretty hazy, that was, 'cause it
was based on a long series of experiments. It had, it had several antecedents, and then,
it's been under development since then, so there, it doesn't really have an end date.
But it's, the Flame algorithm produces images that are organic and fluid. It's the basis
for the electric sheep and the basis for the artwork I have in the show here. It really
-- what's really different about it, compared to, you know -- there's a number of open-source
graphics tool-kits that are available these days. What's really different about the Flame
- and they're based on the metaphor of the pen and the object, just like computer graphics
has been since its inception, because it was -- computer graphics was based on, on drafting,
engineering design, you know, which was, which was done, with, with those objects, and so,
just like - with most new media, the first thing they do is imitate the old media. Like
when movie cameras appeared, the first thing they did was point them at a, at a play, and
then run the theatre, and just record it.
But the Flame algorithm is really, operates by a completely different principle. It's
based on solving an equation with thousands of parameters and millions of variables. It
takes a long time. And I've actually just completed a documentary about it. So, this
is just, it's a three-minute little video, that I'm gonna play.
[Scraping sounds]
[Sounds of the volume being turned up]
FEMALE NARRATOR: In 1993, when the World Wide Web was new, Scott Draves released the Flame
algorithm as open-source. This was, quite possibly, the earliest application of the
GPL to Art. Since then, his code has been copied, expanded, and is still thriving. Every
day, people all over the world create Flames, with no intervention or permission from Draves.
As an example, Flames have been used recently in works by former Prix Ars Electronica prize
winners. "Loka", a music video by Glenn Marshall, released in 2008; "Strange Culture", a movie
about Steve Kurtz by Lynn Hershman Leeson, released in 2007. In fact, one of the very
earliest Flames, Flame Number 149, by Scott Draves, won a Prix Ars Honorary Mention in
1993. Now Flames are used by commercial artists in Films, Books, TV and Advertising, as well
as by thousands of hobbyists.
Flame is not an end-user application. It is an Algorithm and API, but a departure from
the usual Graphics APIs, which are based on the metaphors of the pen geometry in 3D. Flame
images are formed by solving a recursive equation with thousands of parameters and millions
of variables. The results are distinctive, recognizable, and extremely diverse.
How do designers make Flame images? At first, they used Flame plugins, made by other programmers,
for Photoshop and Adobe Aftereffects. Flame is on millions of desktops world-wide, having
shipped standard with LINUX for years in the GIMP. Eventually, more powerful standalone
apps appeared, including Apotheosis in 2004, Oxidizer for Mac in 2006, Cosmic for LINUX
in 2008, Apple 3D in 2009 and the cross-platform Frost in 2009. Following Flame's open-source
GPL licensing, these applications are all also GPL.
The Electric Sheep is another important design source. It's an easy-to-use open-source screensaver
that animates Flame designs and creates new Flames using a popularity-driven genetic algorithm.
All are creative-commons licensed, for remix and re-use. Designs frequently appear in do-it-yourself
music videos, and VJ performance.
[piano music]
Making the decision to open-source one's art is a philosophical statement in support of
creating a better society, one with more creativity, that is more participatory, less prepackaged
and broadcast; an artist who makes open-sourced artwork [piano cuts out] actively relinquishes
control, realizing that other may use it [soft techno beats begin] for questionable ends.
But the artistic rewards Draves has received by sharing the code has been tremendous. It
has been vastly expanded by contributions from all over the world, and totally rewritten
in 2009.
[Beats end]
Flames have become their own genre, currently ranking seventh in Google under the search
term "Flames". Scott Draves has spawned an immense, loosely-joined world-wide community
of designers, programmers, and passionate fans. [sweeping instrumental music eases in]
The open-sourcing of the Flame algorithm and how our culture has responded to it show how
the free flow of information allows an artist to exceed their own boundaries.
[music crescendos]
SCOTT: So...[pause] So the Flame Algorithm is the, the basis for the Electric Sheep and
the basis for Generation 243, which is the, the piece of artwork here in the shop. But
- and it's really, the Flame algorithm, it takes hours to render even one single frame
of, of video, and video, it plays back, when you see it animated at thirty frames per second,
so on a regualr computer, it's a day of work, per second. And so, this kind of artwork is
only possible with thousands of computers, essentially the collective power of the Internet,
all, all working together. And the, the process -- what I'm really interested in is, with
something with open-source art, is not so much making these specific images, but really
allowing other people to make their own images, and enabling [member in the audience coughs]
other people to make their own art, which is kind of the, the participatory, it's the
ultimate form of, of interaction. But I've also done sort of more traditional interactive
art, and interactive art is, is really become popular in the past, say, five years. But
it was [pause] the - The Bomb was an interactive software artwork that I did in the mid-nineties
and it was an interactive visual musical instrument, and what that, what I mean by that is, it
was something that you could play like a musical instrument, like you would play the piano
or the drums or the trumpet, but instead of making sound, it made, it made visuals. And
it used the, the keyboard, the mouse, and audio input, and then it did generative video
out, out, output. And it was based on a whole bunch of techniques from complex systems theory,
in particular cellular automata and reaction diffusion, computed in real-time. And it actually
included versions of the Fuse algorithm and the Flame algorithm, real-time versions of,
of those as well.
So it was just kind of a, it was a big toolbox or a - which is like, which I used for VJ
performance. So you wanna take this to a show, and it was taking audio input from the, from
the, from the DJ, or from the band, and then I was using the Keyboard and the mouse to
sort, to riff off that and to create a video show, which was projected behind the, behind
the band or behind the, the DJ. So, and that's what got me into, into live performance, I
think, back in the mid-nineties. So -- and actually, later on, this, because this was
open-source, again I was able to sort of, things happened to it that I wasn't able to
do myself. So in this case David Zicarelli, who does the Max/MSP program, actually took
Bomb and ported it into Max/MSP so that it was a module, inside of that interactive designing
environment, which has turned out to be really popular and I guess Luke is actually involved
in, in developing that system now. So, so anyways people were able to, other people
were able to perform with this and to create their own music and so -- unfortunately this
software is really old and actually I've, I've let it lapse. It doesn't, it doesn't
run on modern computers any more, but I have a clip which I'll, which I'll play for you.
And so this is a music video. It's sort of like a, a canned VJ performance, and one that
I created using, yeah that Max/MSP System.
So remember this is, so this was, and this was done on a really, I think 486 laptop,
so it's really old PC that you would, you'd, you could barely even boot up today. But it
was computed in real time; it recomputes every pixel on every frame. So I'm just gonna, I'm
gonna skip through a little bit here just to give you an idea [Strong brass note] of
what this program can do. [symphonic sounds].
You can barely see it at all. These projectors are just too dark, I'm sorry. [music continues...
cuts out abruptly] What it's, what it's doing right now is, it's running a reaction diffusion
in real-time, and reaction-diffusion is the algorithm that animals use to create their
spots, essentially, like leopard spots and tiger stripes and stuff like that. So that
what you're sort of seeing is like an animal texture, an animal skin which is reconfiguring
itself in, in real time. [music comes back on. more flowing, dissonant] [fading][builds]
[violent cymbal crash] [swift percussive line][jarring sounds][piano crashing]
Ok, so that's just, that's just a taste of that one. So, and then I got, I got tired,
so that when I was doing, when the, when the, when Bomb runs, it has to generate the graphics
in real-time, which means thirty times per second. You only get thirty, like thirty milliseconds
to create your, your, your image, which is a big constraint. And so I got, I got sort
of tired of that and I wanted to do something which was really high-quality, so I went back
to the Flame algorithm, and this was, oh, I was living in the San Francisco, and the,
you know the Dotcom explosion was happening, this was 1999, and so I did the Internet-distributed
version of the Flame algorithm, which was the Electric Sheep.
Ugh. So, in the Electric Sleep the screensavers form a distributed supercomputer. So when
you're not using your computer, it goes to work creating art, and the, the audience votes,
everybody who's watching that screensaver can vote on whether or not they like what
they see and that runs a genetic algorithm so the, the artwork is evolving, like with
Darwinian evolution, so it's this artificial intelligence which is trying to satisfy human
desire, and there's also crowd-source in the genetic designers, so that people are uploading
completed designs into it, and the human design team collaborates and competes with this AI.
And so this project has been running online for about ten years now, and so it's one of
the first, sort of crowd-source or collaborative Internet artworks and it's still, still going
strong, still gaining speed. Right now we have about 350,000 users monthly uniques.
And here, I'll, I'll show you, you can see what it looks, like, 'cause it's my screensaver.
And, it's, it looks, it's in the same basic style as Generation 243, but there's a couple
of differences. One is that this is really unedited, so there's, because it's on the
Internet and it's responding to the voting public, you know, that means, it, also it's
like, just like democracy. It appeals to the lowest common denominator [laughter].
And so I call it the Las Vegas problem, which is that things that are like bright and flashy
get a lot of votes and reproduce, and then, so you end up with a lot of, a lot of stuff
like that, which is not what I think is really truly beautiful. So the other problem with
it is, while it's great to have 350,000 people sort of helping you create your art, it really,
it, it, it sucks to have that many people downloading videos from your website. Your
server, the server bills are really out-of-control. So, I wanted to - plus, it's low-resolution.
I wanted to increase the quality. So essentially, in order to solve these problems, I started
to create limited-edition artworks, instead of stuff that was just software artwork that's
free online. And that's where the Dreams in High Fidelity series came from, and that's
where Generation 243 came from, the piece here. And so that was sort of going from the
Internet sort of to the, to the museum and to the gallery. And the difference, the sort
of, the things that differentiate it is for one, one thing, in order to make Generation
243, what I do is I go through the thousands of things that the screensaver does, and it's
it's, it has hundreds of thousands of sheep have been produced, and I pick out a few of
them that satisfy my aesthetic, that I think are truly beautiful, and edit them and assemble
them, redo them in high definition, in high quality, and high bandwidth, sort of beyond
the quality that you can do video over the Internet, and then turn it into a limited
edition piece of art, so, something which is collectible, and, and maintainable, and
has a permanence. So, the one we have here was commissioned by Carnegie Mellon University,
and sits in the lobby of their new Computer Science building, so... and there's, and there's
more, more on the way. But I think I'm out of time. Maybe we have time for just a question
or two?
JOSH: Thank you, Scott. [Applause] Does anyone have any questions?
[unmic'd question from audience]
SCOTT: You mean how much do I play with pre-existing algorithms in general? Or how much is that
piece, or is that stuff based on pre-existing ones.
[unmic'd indistinct followup]
SCOTT: Well the stuff that I have done are, they're inspired by natural algorithms, for
example the reaction-diffusion stuff and in general my art, it tends to cross the line
from digital into organic, and sort of take things from, that have natural beauty and,
and recreate their essence in software form. But that's done, it's never done by literally
copying or reproducing the algorithms from nature. It's, which tends to be really hard.
So instead what I do is I, I find something that, that, that has the same effect or captures
the essence of, say, leopard spots, even though the actual equation or the actual algorithm
is, is not derived from studying a leopard, for example. It's, it's just, but it does
have the same, the same overall properties.
AUDIENCE MEMBER #1: One of the things that, that's interesting to me about this business
of getting from the Vegas effect and then you know, sort of going back -
SCOTT: Yeah.
AUDIENCE MEMBER #1: From the point of a view of a mathematician or a programmer, the elegant
solution is the simplest piece of code that'll get you the most complicated phenomena and
at the point at which you're, you're pulling back from the experience of what's on the
screen, do you start to bump into that ambition?
SCOTT: Well, I, I still have the ambition of producing as much variety, as much quality
as I can with as little code as I can. And that's definitely sort of part of, part of
my aesthetic is synergy, or getting out more than you put in. But yeah, definitely, just
maximizing complexity, you end up with randomness, essentially. And so I found that, you, you
have, definitely have to turn it back, and that's done you know, just manually, by my,
by my own hand. I haven't found an algorithm which can encapsulate what I think is beautiful,
and I'm not really -- I'm kind of curious about it, but I'm not really, that's not really
my mission, so I mean, I, I enjoy making those decisions, I'm not motivated to automate that
aspect of the art, of the process.
MIC'D AUDIENCE MEMBER 2: If you were disappointed in the Vegas, what were you, what were you
hoping might come out of such a collective endeavour?
SCOTT: Well, I, I wasn't disappointed, I mean it's just a fact of life, it's like, you can't
get angry at gravity. So - but, but it does produce a lot of great stuff, and the things
that, and it does produce things that I think are truly beautiful. The, the, what - they're
not necessarily, are, are usually not the ones that were the, the winners of the popularity
contest? But they are part of the ecosystem, which is run by the popularity contest. And
so - and I think that's not too different from our world.
MIC'D AUDIENCE MEMBER 2: If I may, if you look up the definition of beauty in the dictionary,
which I had to do a few months ago, you'll find it says, "That which reveals to the mind."
MIC'D AUDIENCE MEMBER 2: What is missing in that definition? I mean you use the word...
SCOTT: Well, I mean, it's more than just the mind, I mean, to me, beauty has a visual aspect
to it? Or a, or a sort of an aesthetic aspect, which is from some separate from, [loud audience
cough] from, from practical [loud cough again] practical or utilitarian purpose, so there's
a lot of things that appeal to me for practical reasons that I don't necessarily think are
JOSH: Let's move on.
SCOTT: Ok. We have...
JOSH: Scott will be, will be around when the, when we're done with our session's second
presentation. So, feel free to approach him with questions. Thank you very much.
[Loud audience applause]
JOSH: Our second speaker this evening is Luke DuBois. Luke is a composer, a musician, a
performer, artist, multi-instrumentalist, all-around nice guy. He is also a fairly serious
computer scientist in his own right, and he is represented in our -
LUKE: I would dispute that.
[audience laughter]
JOSH: Hacker?
LUKE: No, no, no. Definitely not a hacker.
LUKE: More like a comic hacker.
JOSH: Those are the best ones. He's represented in the, in our exhibit, by his work [audience
member coughs] "Hindsight Is Always 20/20", which is a fascinating visualization of presidential
inaugural speeches, that let's you, that captures a moment of history in a very interesting
way. And... thank you.
LUKE: Alright. Thanks, Josh. So, I'm gonna give you the whirlwind tour of my life, in
30 minutes, including some older work and a couple things that I'm working on now that
I haven't quite figured out what they're gonna look like or what they're gonna be. But, and
I'm gonna focus on, like Josh said, I do, I do a lot of things, and I'm gonna focus
on, mostly on, I guess you, what you would call my visual art, stuff that I do that has
some sort of visual output. It's meant for viewing outside of the context of performance.
But to give you a little bit of background. I started out as a, well, as a musician, and
actually, before I even started out as a musician, I started out as an engineer, at Columbia
University. And I basically flunked out of engineering school because I was too busy
being a musician. And I was working at a place called the Computer Music Center up at Columbia,
where I was restoring and performing music on all these old analog synthesizers from
the '60s and '70s. They looked like phone switchboards. And I put a band together in
the nineties called the Freight Elevator Quartet. This is us at the kitchen in 1999, and there
were four of us, we had a cellist, and a didjeridu player and a guy with a bunch of drum machines,
and I used to play these old synthesizers. And as you can see, if you sort of focus on
this bit of the image, and ignore everything else, we're really boring to watch. This is
the conclusion that we drew, after, after much consideration. And that was because,
well there was, for a few reasons, but I, I would argue a big chunk of, of what made
us not that interesting to watch was that we were engrossed in a lot of fairly obtuse
technology onstage that didn't really have a lot of, I guess, sort of, what people who're
smarter than me call "performative agency". That's just a fancy way of saying, when I
do something onstage, you can't tell what I'm doing, and how it affects whatever you're
And so, so the, the, the easiest way to do this is to compare for example the cellist
and myself. That's me with a laptop computer. You know, when a cellist plays a note, you
kind of know what's happening. Alright. You can kind of expect there's a, there's a physical
motion with an acoustic effect, whereas me, I could, I could be checking my email. [audience
giggles] I wasn't I don't think, but, but I've done that before, absolutely, [more audience
response] checked my email onstage, and, and you know, anybody who performs onstage with
a laptop and they say they've never done that, they're lying! [Audience laughs] Right? So
that's just the way it is. And that's okay. But, but a lot of what this early stuff I
was doing was, was trying to grapple with a couple problems. One was the idea of having
a computer onstage. And it's not that a computer is technology. So this is actually an important
sort of thread in some of the ideas that I have spinning around in my head at all times.
It's not that this is the first use of technology onstage. Music is technology, unless you are
singing unamplified in a forest, and the music you're singing is something you made up in
your head that morning, you're using technology, right? If you're singing something that was
ever notated, you're using literacy. If you're singing in a space, you're using architecture.
If you're playing an instrument, that's technology. If it's amplified, that's definitely technology.
Right, if it's amplified, and produced on a record that someone plays later, that's
definitely technology, right? So this is just one more step in the equation, and to say
that this is substantively different is actually to a certain degree a moral panic. Right,
it's sort of, actually, it's not, right? And when we talk about information art, which
-- I'll show you a few things while I'm showing things on the computer -- that's another kind
of thing that is interesting that I did, information art, sort of might have a startpoint sometime
in the 1970s, when I would argue that, you know, we've had information art for millions
and millions of years -- we just call it music. Music is the sonification of data, for all
intents and purposes. It's a bunch of constructions, being laid out in time, and effected or evoked
in a performance.
So, so I was trying to figure out how to make this look more fun [laughs] --back to the
point. And I started performing and writing a lot of music that involved imagery. And
so this is an early example [strumming audio starts] This is a piece from 2003. This is
a piece for guitar, it should be obvious at this point that I do not have quite the sophisticated
visual sensibility of Scott. This is called Turtle Graphics, I don't know if you guys
remember this. This is for a guitarist, guy named Dominic Frasca, and the different notes
on the computer, the different notes that he plays are, are telling the computer to
do different drawing actions, drop forward, turn left, turn right, and if he plays it
perfectly... [silence] he gets a plant. [Audience response] [music resumes] And if he screws
up, he gets a different plant. It's called a mutation. It's how we got here. It's not
a big deal. And so you can hear there's a sort of a riff that... a very riff – [hums
along with guitar music]. There's a melody. That melody is the graphical primitive of
this plant. OK, see, instructions say, it's drawing a little branch and going back and
forth over the original path, happens like 8,000 times. [music ends] And the whole idea
of this was that I was trying to deal with the issue of sort of, music and modern music,
and in particularly modern music that has an algorithmic basis to it, which is increasingly
common, where instead of sitting at the piano and writing things that sound beautiful, you
actually can see part of the compositional process doing algorithmic piece of math. Right,
and how do you make that transparent to the audience? How do you make the system at play
actually make sense to someone who's listening to it, so it doesn't just sound like noise?
Well one idea is to use systems that have a linguistic apparatus to it, which is what
these algorithms -- they're called Lindenmayer systems – sort of do. They can be broken
down into hierarchies, the structure, which is how language works, and is how we listen
to music. And so I started thinking about that, and I started thinking about that is
a useful and kind of fun way to grapple with making sense of information, which is to transcode
it, stick it in a medium you don't usually experience it. Just to give you a couple [audience
coughs] quickie examples from the repetoire, this is another performance algorithm that
I use once in a while. And of course, it's not working. Hello? Oh, there it is! Oh, it
is working. This is a, these guys are listening to me, right? Bleh! [audience titters] Right?
They're they're, listening to me, and they're visualizing my voice. And this is something
that I traditionally would do with a, in a performance situation, with a friend of mine,
a guy named Todd Reynolds, who's a really amazing incredible violinist. And the conceit
of it is, you have these four birds flocking around, and what they're trying desperately
to do is detect my rhythm, and stand there in canon based on the rhythms of my - I'm
just talking to you, so it's not quite working. But it gives you an idea of - the idea that
you can, you can visualize a sound, in a, in a, in an interesting and kind of kinetic
way. You can also go the other way. And so this is... just to to give you this little
example of this, I'm gonna... while I was working on all this music, I helped to write
a totally wacky programming language called Jitter, which is what this is. I'm gonna talk
into my computer for a second. "My dog has fleas! Testing 1, 2, 3..." Ok,
let's see what that looks like that. Mmmm that should...
"Testing 1, 2, 3" There you go! So this is a, this is a sonic match -
"My dog has fleas! Testing 1, 2, 3..." Can you all see that? This is like a pretty
standard way to turn a sound into a picture? Right? The red line's time, and the left to
right is the frequencies, and the height is sort of how loud a frequency is at that point
in time? That make sense to you guys? Alright. It's a pretty normal way to -
"My dog has fleas! Testing 1, 2, 3..." So the dirty little secret here though is,
the sound you're hearing - "My dog has fleas! Testing 1, 2, 3..."
is being scanned off the image. It didn't actually record my sound, it just turned it
into a picture. And now it's playing back the picture. And so you can do these things
that you're not supposed to be able to do with sound. For example, there is absolutely
no reason why the red line has to move. "My dog ha-" [words turn to ambient sound]
[indistinct words, and then the words slowly resume out of the ambient sound]or to make
it run kinda slow - or backwards [the words begin to play in reverse] or make it run really
fast [the audio becomes a high looping rhythm], right? But like really, really fast. [high
rhythm simply becomes an even high tone]. Turn it into an oscillator.
You can do that, or you can say, you know what? It's a picture. What the hell, I just
blur it.
And then... [tone acquires a lower, wavelike pattern] then when you turn that back to normal
speed - [echo-y, wandering pitch to his original words] You blur pictures all the time. Why
don't you blur sound? That's what you get.
So I started thinking that was pretty cool. The idea that one of the neat tricks about
turning a sound into an image, is that you could essentially divorce it from time. And
so I started thinking to myself, well what kind of things would, would I wanna do with
the idea that the red line can move at any speed, and also that you can blur things.
You can basically create sort of, almost sort of long-exposure photographs of sound, and,
and I got into this whole riff on this technique which I started calling Time-Lapse Phonography.
And I started thinking of like, things to put through it, so here's all the billboard
#1 songs. Sped up to 1 sec for every week they're at #1. And so you got um [resonant
jarring sounds] and in [low, frequent hum] 37 minutes the entire history of the pop chart
done as a sequence of drums. [Rhythm and frequency rises and falls] Right? And some of these
drums are interesting, right? Like, there're some things about the drums are interesting,
right, so that like all of 1978's in the key of F. And... [low hum] Beegees songs... Alright
[sound ends] and if you listen to the, and if you listen to this stuff, right, you have,
you discover pretty quickly that you have part of this memorized, right? If you grew
up listening to pop music, [music rises from inaudible again] you see it sort of going
by and you're like, "Oh my God, I remember that song," or "Oh I hated that song" or,
"I lost my virginity right there," or you know you have the whole thing like in your
brain, right? And I started thinking about it as, as, as a aspect of our culture, and
me... I'm kind of obsessed with lists. The idea of top lists, or #1 lists, so I started
making a bunch of pieces around it. So, along with Billboard I did this. This is every Academy
Award Best Picture sped up to one minute. Right, so this is 76 movies in 76 minutes.
This is "Wings". Let me find you something you all might recognize. That was "Gone With
the Wind". That was "Casablanca". "Casablanca". Right? "Casablanca". The entire movie's in
there. It's an average, ok. It's skipping through the film alright, [sound swells] so
it's a weighted average. Musical [sound ends] quite beautiful, actually. Let's see if I
can find a good music --, like "West Side Story" or something. Aw, you guys get the
idea, but um - I'm too late now. Gotta go back a little. Yeah.
So the entire [ sound rises and swells.] There we go. So you've got entire films in a minute...
no matter how long. You have to write them down [sound swells].[abrupt ending]
I made a piece, where I took all the Playboy Playmates of the Month, and centered their
eyes. So this is fifty years of Playboy in fifty seconds, alright so this is time-lapse
pornography [audience laughs]. Right.
And so I started thinking to myself, well you know what, like, we got these lists, and
they're important to us, but we don't necessarily think about like how they got to be those
lists, right? You know, it's, it's, it's, it's a little bit like trying to actually
debug how a search engine might work. Like how did, how did #1 songs become #1 songs?
But to have a #1 song you need a collusion of several major multinational corporations,
right? To get a, to get a, Academy Award Best Picture, the people who vote for the Academy
Awards are the people who are themselves eligible to win at the Academy Awards, right? We did
not vote for "West Side Story", right? It's a peer review. That's why they thank the Academy,
right? And don't even get me started on Playboy. But you get the idea, [audience chuckles]
right? We did not select these things, right? We live in a democratic society in which the
most valuable cultural objects in our society were not democractically selected. Alright,
so maybe that's kind of interesting. So, I started thinking about that, and I started
thinking about other kinds of data sets that you could subject to that kind of analysis.
And I'll, and I'll show you just briefly, the piece that's on the wall, around the corner,
I was, I was commissioned by the, the Democratic National Convention to make a piece in 2008,
and, and I made a piece called "Hindsight Is 20/20". And, and I knew, I knew I was gonna
make a thing about presidential something, presidential speeches, presidential rhetoric,
something like that. And I, I started out, thinking I was gonna do something kinda like
a billboard pop chart, like, hundred most common phrases or something like that, you
know, so like, you know, George Washington, you know, the state of the union is sound
or whatever. And then I saw James Carville on CNN, and he was on some rant. He was talking
about vision, and he's saying, "Well the - " he was talking about the Bush adminisitration,
this was back in like 2006 – he's saying, "Well the problem is they don't have vision..."
And I started thinking to myself, you know, how do you judge your leaders? You judge your
leaders based on their vision. So how do you test their vision? You make eye-charts. So
I started to make a bunch of eye-charts, one for every President of the United Statess.
These are the words that their State of the Union addresses most common / least common
words. So George Washington, his most common word is "gentlemen". George Bush is "Terror".
Some of these you can guess, right? Ronald Reagan is "deficits". Richard Nixon interestingly
is "Truly". Some of these you can probably figure out. Like Abraham Lincoln is "Emancipation",
right, something like, FDR, FDR is "Democratic", right, Herbert Hoover is "Unemployment", and
Calvin Coolidge: this is how you get veterans, and this is the last time New Orleans exploded,
right? And so what do you do with this? Well, what
you do is you make yourself 43 two-scale massive light boxes that actually work as eye-charts.
If you stand 20 feet back and you can read between those lines you have 20/20 vision.
You install them outside the Pepsi Center in Denver, Colorado, during the Democratic
National Convention. Lots of people walk by, it's very fun. And you, and you also make
some prints. These're prints that, you know, we made that, were shown at the Weisman Art
Museum in Minneapolis during the Republican Convention.
And they then sort of trail around, and I thought this was actually kind of interesting.
This is the National Constitution Center in Philadelphia. This was where this was during
the actual election in 2008. And what's, what's interesting about them is, you get this really
interesting insight into Presidential rhetoric of any given time period. And the idea behind
the piece was that the State of the Union address is a piece of political theater. It's
a Constitutionally-mandated piece of political theater. It's actually required by law, right?
A President has to come, do his homework. And what it is is it's about a sovereignty
relationship It's about how Congress is sovereign over President. The President has to report
to Congress, not the other way around. The King cannot dissolve Parliament -- Parliament
can impeach the King. Right? So that's an important power relationship that's embodied
in this, actually in this event and at the time I was working on this, we were living
in what will arguably in hindsight, we felt was an Imperial Presidency. So I wanted to
do a big piece on how Nancy Pelosi was really George Bush's boss, [deep sniff] and how we
had forgotten about that. At the same time, so let me just show you
a few other little bits and pieces, at the same time I was wrapping up, at the same time
I was doing pre-production on Hindsight, I was wrapping up a film project that was a
collaboration with my dear friend [sound rises] Lian Amaris Sifuentes, who's a performance
artist, and it's, it's a kooky collaboration, it's a three-day-long performance, scripted,
directed, and executed by Lian but then I film? Sort of document it, panoptically, like
on a, in as ridiculously obsessed way as humanly possible. Got all these hard drives, and then
I, and then I took the film and accelerated it. So it's a 72-hour performance, set on
a, on a, on a traffic island off the coast of Union Square, as we set up a boudoir, and
she spends three whole days out there getting ready for a night out on the town, doing all
the things that one would do in such a situation. And then I speed the whole thing up to 60-speed
so in experiencing the film, you get a very different sense of the performance. So it's
not a documented performance, it's sort of more of a reaction piece to the performance,
where it's sort of, in a weird way, it looks like she's moving normally and everything
else is flying by. And they scored it to a violin concerto, in seven movements, that
are all about different types and stages of obsessive love, time and relationships, and,
and it's funny, like I don't really think about what I do is, is, is information art,
per se, but this is, this is maybe a different spin on it. So instead of looking at lists
of things and doing a bunch of math, you can also be a lot more lyrical with it and just
say, "I wanna do a thing on the subject of how you look at someone, when you love them,
and how that works, and what there's, what that shop list might be, and what those different
phases are." And then you can just sort of construct something out of it. So a couple
things that I've got kicking around on my computer these days. Let me just show you
one last thing. I'm workin' on kind of a monster of a piece. It's also about love, sort of.
I decided to make my own census, the census year, it's 2008. I decided to make my own
census, so what I do is I joined 21 different online dating sites [audience reacts with
amusement] as a straight man, a gay man, a straight woman, and a gay woman, and downloaded
everybody, so I downloaded 16.7 million Americans to a hard drive. And sorted them by congressional
district. [audience laughs] So
I can tell you how many people in your district are shy, and how many people are looking for
someone who's fun. Here, I'll, I'll show you [loud audience laughter] So, for example...
there ya go! Here we go, here's some math. Here are all the people, here are all the
shy people in the United States. If you're purple, it means you're more shy; if you're
green, that means you're less shy. So red is shy women, blue is shy men, ok? Here are
all the lonely people. White means a lonelier district, right? Here's where people say they're
sexy. [audience laughs] You know, so sexy really only shows up a few places. It's actually,
it's actually pretty interesting. It sort of shows up on the coast. Oh and here's where people say
they're funny, men versus women. So if you're, if you're-- the white areas are the least
funny parts of America. [audience laughs] God. Right there, the least funny part of
America. Right there. So I'm preparing a whole bunch of maps, and
I'm preparing a census document, that I think actually is vital to our national security,
because what it, what it, what it allows you to do is it allows you to prepare maps and
other documents. So this is sort of thing of -- here are all the single people in New
York, with their eyes centered. But it, it allows you to prepare a pretty substantive
case for critiquing or debunking the censuses that exist now as a snapshot of, of, of us,
as Americans, right? Saying that what we should really be paying attention to is media and
household income, and how many of us are in manufacturing sector jobs, and how many of
us are in service industry jobs, and how many of us have changed jobs. So, so we do this,
and we make these chloraprep maps so it's kind of white to black maps about that stuff.
And you know, they have meetings on it, they, and then, policy is actually affected on this.
So what if instead of that, what if instead of that we did, we did red state, blue state
maps, based on type A, type B personality, right? Or, what would happen if, if I told
you that, rather than median household income, we just sort of looked instead at, you know,
how many people are lonely at night? And where those people are, right? Or if we looked at
-- just to make it a little bit more funny, like I joined, I joined 21 sites, I joined
a lot of alternative-lifestyle sites, so I can tell you for a fact that men in the Eastern
half of Long Island are way more interested in being spanked than men in the Western half
of Long Island. [Audience laughs] Alright? And I think, I think Congress needs to know
this information, I think this is important, this is important stuff. So, so this is, you
know, so this is my, this is my current, this is my current - it's, if you haven't figured
out yet that I have obsessive-compulsive disorder, you weren't paying attention. But like, but
that, but that's sort of the idea, is to, is to make this body of work around that.
And it's called "A More Perfect Union", and I'll be showing it sometime in January.
We good? Is that about right? Did I come in around time?
JOSH: Yeah!
LUKE: Alright.
JOSH: Yeah.
LUKE: Alright, thanks. [audience applauds]
JOSH: If nobody has any questions after that, I'll be disappointed with you[audience chuckles].
LUKE: Ask me a question, anybody.
AUDIENCE MEMBER #3: How long did it take to do all those compilations?
LUKE: What, like, which compilations?
AUDIENCE MEMBER #3: Oh, like, for instance, the... the movie one.
LUKE: Oh, jeeze, that took a long time. I was actually really lucky, because I did that
piece like right before all the video rental stores closed? Right, like when they were
still like three kims? And they were still like the movie place on the upper West side
and stuff. So it's, so I was really lucky because I didn't, I didn't have to, you know,
actually let, I could, I could, I could run around town and go into movie stores and be
like, "Do you have Grand Hotel?" And they would be like, "Sure." And like, I was able
to acquire all the, all the movies, and then I had a render farm, a lab of computers at
work where I was able to crunch them so that took about four months. Four or five months.
Yeah. Took about 16 hours a movie, so, but that was a while ago. Now it would, wouldn't
take that long. Yeah. Some of the things that I do. The Census I've been working on for
about two years, so it's, you know, takes a while, to crunch all that stuff. Sir.
AUDIENCE MEMBER #4: Just curious, it, you seem to have to work between the lines...
LUKE: Yeah.
AUDIENCE MEMBER #4: Lining up the eyes, that seems very very difficult -
LUKE: Yeah, yeah. It's a drag, [audience laughs] I mean you can automate some of that, right?
AUDIENCE MEMBER #4: Well, that's my question.
LUKE: Yeah, you can automate some of that. Your, your head is, most people's head, within
ten percent is, is five by seven in the units of width to your eye. So, so, if you can find
a face, you can center on. It's like you can train things to do that. Problem is in online
dating profiles, people get a little arty with their photographs, and so, yeah, so it's
like, so it's like, a little bit harder. It's much easier to do like, oh my God, what's
his name, Jason Salavon, digital artist in Chicago, has a lot of pieces with things like
graduation photographs, where he does that, right. That's much easier to do because there's
like this sort of guarenteed neutral background, right. So that's, that's a sort of automatable
thing, maybe? It's, it's, it's a little bit harder to do with this, so I've been spending
a lot of time clicking on them.
I'm working on a computer vision piece right now that is all about trying to automate things
which is -- I've been making a bunch of pop icons. So, what I did is, I, I trained the
computer with a thousand paparazzi photos of Britney Spears, and then just run all her
music videos through it, and it can lock her eyes. So this is, this is an automated process,
this is, finding her eyes. But that's not, that's, that's not like super-duper easy to
do. Yeah, so. What else?
AUDIENCE MEMBER #5: Now all the Presidential Speeches? How long did it take you to figure
all that?
LUKE: That wasn't, that didn't, that, that wasn't so bad, because all that – all those
speeches are a matter of public record, so it's, you just have to find them and there's
a, there's a wonderful place at UC Santa Barbara called the American Presidency Project, it's
part of the Pop Political Science department, where they've been taking every, basically
the entire paper trail of the American Presidency, every speech, every bill they signed, every
judge they appointed, every whatever, and sticking into a huge database for researchers,
and they've been doing it for about ten years. And so they have a pretty definitive set of
tools and access points if you're doing research on the Presidency, where they have a whole
like, it's, it's called a MySQL database, they have one of those things where you can
just get 'Time' on the thing, and look for stuff and do word searches. That's how I made
these, was I, I basically unitlaterally announced myself as their first Artist in Residence,
without them knowing what the heck that meant, [audience chuckles] and just sorta, just sorta
invited myself to use their database, and they were like, and they were like: "Sure.
That's fine." So that's, that's where that came from, is, is -- so that wasn't super-bad.
The, the biggest part of that was coming up with the look and coming up with the algorithm
by which I was trying to get at the right words, 'cause actually all the words in the
eye charts are unique, so not only is it like the top 66 words for every President. But
it's also, if, if the President has that word in their chart, that means it's not in any
of the other charts, it cancels, right, so George Washington is "Gentlemen", that means
not only did he say gentlemen the most, that means he also said "Gentlemen" more than any
other President, right? So George Bush is "Terror", that means not only did he say "terror",
"terror" the most, hermetically in his own speeches, but he used that word more than
anyone else. So it's gone from their charts. So up comes the rest. And then -- so you end
up with these very, very specific historical references to different times to different
things. Yeah? It's like a tag club from hell. Um, yeah, yeah. What else? John?
John: So, Luke, has your personal sense of time gotten elastic somehow over the years?
You know, you're -
LUKE: Ah, jeeze, that, you know in hind, well, yeah, I mean dude, given what I used to do
to my body in high school, I don't think this matters but it's like, I don't know, maybe,
I definitely have started taking the long view of how we deal with technology in our
culture, and how we deal with information in our culture. Like I think one interesting
little thing, I had this huge writer's block moment a couple weeks ago. I was trying to
make a blog entry, for the Guggenheim, for that YouTube Play thing they're doing, that,
that video art by mail that they're doing with, with [indistinct] Guggenheim and Youtube
asked me to be a blogger for it, and I was trying to come up with some introductory post.
And I was having one of these like crisis of like Online Publishing Video Art. And I,
and I sort of came up with the idea that, -- or, or, just it was, I started thinking,
I started thinking like, ok, so computer programming, right? Like a thousand years from now, everybody's
gonna program computers. It's just gonna be life, that's just the way it is. It's gonna
be normal, right? Doesn't mean it's gonna be easy, it's gonna be normal. Right? A thousand
years from now, everybody is going to publish their work online, right? Right, so it's,
so it's, it's the, it's the Warhol deconstruction, instead of everyone being famous for 15 minutes,
everybody's going to be famous for 15 people. Right? That's just the way it's gonna be.
Right? And so looking at, looking at this particular slice of time and our world and,
and sort of doing things like bemoaning the collapse of the record industry -- so the
record industry will just be a historical fifty-year blip in the radar, the idea that
people should make money off their records, in the grand scheme of human civilization
is gonna seem so like, random and pointless, right? Once we're a thousand years out or
what have you. So I have started kind of rustling up some ideas about that, but my own experience
of time is that I never have enough of it, and that's been true, that's been true for
a very long time. Yeah. What else, folks?
AUDIENCE MEMBER #7: So is that why you speak so fast? [audience laughs]
LUKE: Know, I think, I think really fast because I - I'm always, well, yeah, actually, that
might be why I speak fast. I also, I also have this terrible accent, I'm from Hoboken,
New Jersey, so I don't have a, I don't have a very good, I don't have a particularly palatable
accent. So I guess if I rush through things that's the best way to, to get it over with.
[Audience laughs] Yeah. See, you can, can listen to somebody who, who, who has a better
command of spoken English. But yeah, it might be why. What else, guys? Sir. Kurt! Oh my
God, how - how are you?
KURT: Do you think it's like to, [indistinct questioning] but isn't it possible that there
are actually these active listeners [indistinct ending]
LUKE: Yeah. Oh yeah, yeah. I had a meeting with somebody a couple weeks ago who seemed
extremely interested in the fact that I made some wacky algorithm to sift all that stuff
and yeah. Yeah, absolutely, I mean, there, there's a whole body of sort of computational
research right now in everything targeted advertising to whatnot. It's, it's about making,
trying to make sense of things like that, online dating, profiles and whatever. I mean,
my piece is, isn't really so much about it. My piece is more like the census, like the
actual formal US Census that we do, and what's wrong with it, and how we can like sort of
play around with it, than about the fact that I'm quite, quite honestly -- I'm, I'm violating
the privacy of 20% of the adult population of the United States [audience giggles] by
downloading and, and, and crunching this data, right? I've just, I've committed 16.7 million
counts of, of wire fraud, right? You're not allowed to download online dating profiles,
right, and use them for anything, other than entertainment, right? So, but everybody does
it. I'm sure there's someone at Nielsen right now who's doing it. I'm sure there's probably
someone in your company who does it. [audience laughs] There's people all over the place
who do it. It's just out there, right? So the thing is just to, to make something interesting
out of it, and try to make some, make some sense out of just coming up with something
that is thought-provoking? I don't - I don't, I try to, I try to, if, if possible, to under-editorialize,
in what I do. So like the eye charts, they, they're not, that wasn't a partisan piece.
That was just the way the map turned out, right, and so I showed it at both the Democratic
and the Republican National Convention, and I spoke to a lot of people on sort of both
sides of the political line in, during that election cycle, and you know when you show
them George Bush's eye chart, it begins with "terror", someone on the political left in
this country will see that as, as as a sort of self-evident damnation of that administration
whereas people on the, on the right of the political spectrum will be like, "Yeah, damn
right, 'course it's terror." Right? So, so it's not like, it's not like I have to point
that out to you, or anything like that. It just is. And you know people can draw their
own conclusions, I think. So I don't know. What else?
JOSH: Is there a last question?
AUDIENCE MEMBER #6: Sort of go back to the privacy slash legal aspects of your work collection,
like your map one?
LUKE: Uh-huh. Yeah.
AUDIENCE MEMBER #6: So I think it's, it's fine to have like aggregates of all this stuff
like how many shy people there are, and whatever-
LUKE: Oh sure -
AUDIENCE MEMBER #6: But you're sitting on a database of like, 18 million people, and
- how legal is that?
LUKE: Yeah, I mean, that's, that's, that's the thing. This, a map, a map of shyness,
is not, is not illegal,
AUDIENCE MEMBER #6: Right, but aggregates are fine, but -
LUKE: Absolutely.
AUDIENCE MEMBER #6: But you're in possession of -
LUKE: Right, it's the fact that I own a hard drive.
AUDIENCE MEMBER #6: Yes. [audience laughs]
LUKE: That's maybe a problem. That's maybe problematic. Yeah, absolutely, yeah I should
clarify that. Yeah this is, this is no big deal. Once you get to -- from Point A to Point
Z, the final piece, right, won't be that way. The problem is that, I've got all this stuff.
But I'll throw out all the stuff; I mean I did, I did this map, and I haven't honestly
even read most of the profiles. That would take a while. [audience laughs] Although I
did do this sanity check where I went through to try to get rid of duplicates, and I found
one person who was also on all of them, [audience laughs] right? I found another, I found, I
found another, yeah, I found another faker, like Helena Bonham Carter in "Fight Club".
So I, so I, so I wrote to her, and I was like, "J'accuse!" And, and she turned out to be
a pop singer, a singer/songwriter in Oakland, California, who's doing a whole album where
the lyrics are from online dating sites, right? So I asked her out, right, [audience laughs
loudly] I was like, "Oh, we should go like, we should go out on a date." So, so I flew,
I flew to Berkeley, actually on Election Day Weekend - this is like, a total wacky whatever,
it's like two years ago - on Election Day Weekend, and we went for coffee in Berkeley,
and hung out, and we made up this whole - we didn't hit it off. But we made up this
whole like awesome, like lie we're gonna tell to all our friends that we had this whirlwind
romance that we never, we never - you know. We never actually bothered lying to everybody
about it. But it was, it was, but, she was also on all of them.
JOSH: You actually met through an online dating service.
LUKE: And we met through, we met through 16 out of the 21 online dating services [audience
laughs] Simultaneously. Yeah, yeah. So it was, so it's pretty good. Mm. But yeah, so
you know.
JOSH: This is a massively inefficient way to meet girls.
LUKE: Yeah, I would argue that this is an incredibly inefficient way to meet girls,
but yeah. Although, that's, that's one of things that I always think is interesting
about online dating, right, is that it's, it's this totally bizarre act of self-identity,
where you simultaneously have to put your good, best foot forward. So you have to lie,
and at the same time, you have to say who you wanna be with, which means you have to
be as incredibly truthful as possible to get what you want, right? So, so it's a split-personality
exercise, right? So it's a, it's a, it's an interesting, it's an interesting prose assignment
to have to go through, writing an online dating profile. I did not bother to write any. I,
I just wrote some snarky little thing, but - the, the actual ones I've read are actually,
they're really interesting how people have to construct these narratives of who they
are and who they want and how that works, and so that's what I'm trying to like tear
apart, and get, get to in a sense, so it'd be kind of fun.
JOSH: Thank you very much. [loud audience applause]
LUKE: Thanks man! Thanks. Alrighty. [applause softens]
JOSH: We have a little bit of time left. [sounds of audience rising, chatting] Please feel
free to have a look at the exhibits, chat with the artists. Thank you very much for