Using AI as a creative tool with artist and stage designer Es Devlin | WIRED Live
Released on 03/08/2021
So our next speaker this morning is renowned artist
and stage designer, Es Devlin.
Es is known for creating large scale performative sculptures
and environments that fuse music, language, and light.
She's conceived stage sculptures for Beyonce,
Billie Eilish, The Weekend and U2.
And her practice was the subject
of a Netflix documentary series Abstract, the Art of Design.
She's been named the artistic director
of the 2021 London Design Biennale.
Es wasn't able to join us for this live broadcast,
but wide features editor Vicki Turk recorded a conversation
with Es a few days ago to discuss her approach
to creativity, AI as a creative tool
and the future of live experiences.
Let's see what they talked about.
Please welcome Es Devlin in conversation
with WIRED Vicki Turk.
Ed, thanks so much for joining us.
Thank you so much for having me, Vicki.
I'm excited to be here.
So you make works that can take the form
of huge sculptures or environments,
blending, architecture, sound, light, performance.
How do you use technology in your work?
Do you know what's quite interesting,
is the technology tends to be the last thing on my mind.
Normally what I'm trying to do is tell a story
or express an idea
or try to change people's minds about something
or try to shift their perspective about something.
So I will tend to begin with the shape of an idea or a story
or the rhythm of a story.
Then I'll think about the time, how the time should be spent
by the performers and by the audience
or by the sculpted object in the audience
who might perceive it.
And I'll think about the geometry, the shape of it,
the architecture of it, and the light and the dark of it.
And finally, I'll want to find a piece of technology
that can do what I need to do.
For example, not long ago I had an idea
that I wanted one and a half thousand people
to co compose a poem.
And of course, I wanted magic was what I really wanted.
Normally what I want is not technology but magic.
And I said, well, I want everyone to be able
to donate words and I want a poem to happen.
I want magic to turn these donated one and a half thousand
words into a poem.
And so normally it's in my search for magic,
that very excellent people are introduced to me who turn out
to have been engaged in this type of magic slash technology.
And that's how I arrive at the technology I use.
You know, so it it's always to fulfill a need.
So that's your poem portraits work,
which was using an algorithm to create poems out
of those words donated by people.
Is that the first time you've used ai?
Do you know, it is really, I mean,
I say that I think we are all using AI, aren't we?
Every day and every minute and we are being used by AI.
So I doubt it is the first time I've used AI at all.
But it's the first time I tried
to have some agency in my use of AI, I would say.
And it was with an extraordinary creative
technologist called Ross Goodwin.
And later another phenomenal creative technologist called
Kyle McDonald, who are now both heroes of mine
and what they've taught me about how the algorithms
that can generate poetry through machine learning work.
And I guess the biggest piece of learning, you know,
to me at the beginning, 'cause I was really looking
for a solution to a problem, I was delighted that, you know,
poems emerged in the first instance.
The algorithm was trained on 19th century Victorian poetry,
and the poems that were created by it were rather beautiful.
They were rather meditative.
And an audience's reaction
to them was something like an audience's reaction
to a fortune cookie.
You know, you want to find meaning in it, right?
So it was more reflective on the human species desire
to find meaning than the ability of the algorithm
to imbue something with meaning.
I think in the end, that version,
however, this was for the Serpentine Gallery in 2016,
subsequently, we've been working on this algorithm now
for four years and training it and retraining it
and using more advanced technology
behind the actual algorithm itself.
And the one that we're preparing now for the
UK Pavilion at the World Expo
is remarkably more advanced in terms
of the quality of its output.
It's just remarkably more complex.
But what I've really learnt is,
I hadn't taken into account the bias involved
in the algorithm.
It was one afternoon I was on a call with Kyle McDonald
and he, it was right in the middle
of the Black Lives Matter moment when it had just begun.
And he had been talking to his parents
who were quite traditional in their approach
to things in America.
And he had become very disillusioned with our work.
And he said, Es, you must stop talking about this algorithm
as if it's kind of nebulous.
You must give it an identity, you must give it a biography.
Because to not do that is to deny the fact that each
of these algorithms comes from a specific group of people.
And albeit the UK Pavilion album has been taught by a number
of parents who come from a diverse range
of backgrounds purposefully chosen.
So, however, I think he's quite right
that we will serve ourselves well if we write biographies
for our algorithms
and recognize that each one has its specific viewpoint.
That's really interesting, you know, we hear a lot about
AI and creativity.
You know, can AI truly be creative?
How do we use it?
Is it just a tool?
How do you approach it as a creative tool?
I mean, do you see it as a tool or a co-creator,
or what is the role of artificial intelligence in the
artistic sphere, do you think?
Listen, I think that that question will be answered
in many, many, and broader and broader ways.
There certainly is no single approach,
but in my own experience, like I said,
I didn't set out and say, Hey,
I wanna explore working with AI.
That wasn't what happened.
It was more that I was frustrated
with the fact that one and a half thousand people
could be together,
and these particular one and a half thousand people
who attend this particular evening,
gathering at the Serpentine,
temporarily extinct, I might add.
But for years or mid-June, many of my heroes
would gather, you know,
it would be Zaha Hadid, David Ajay, John Pawson, Ron Arrad,
you know, the designers, the artists
would all converge to help support that gallery.
And I wanted to know, was there a way that a free piece
of art that everyone could take home
for free could be generated by this entire collective?
I guess I was looking for a really smooth
and fluid way for a bunch of humans to collaborate, right?
Just by the pure act of donation
of a word, just giving a word.
So that's what I was looking for.
And I wouldn't have thought of machine learning,
I wouldn't have thought of an algorithm,
but it was really Ross Goodwin who introduced me
to the work he'd been doing for years.
And I guess that then opened my mind to
a collaboration that might be possible
between the living and the dead.
So the poets that we were collaborating with were long dead.
They were poets of the 19th century
who had written these words.
And by training the algorithm, you know,
you know it works more or less like predictive text.
You know, it's consuming so many
of these phrases that it begins to predict
what word is gonna come after the next one.
And that's really the basis of that simple algorithm.
But it seemed to me that I was actually entering to
the ways that our minds worked as humans 150 years ago.
So it was a communication between the living
and the dead, which I thought was quite interesting
between the past and the future.
And tell us a bit more about your creative process.
I mean, where do you get those ideas from
and how do you realize them?
So many of your projects are so huge in scope
in terms of spanning time and space.
I mean, where do you start?
Well, it's interesting because there's been a bit
of a shift in my practice.
And like so many things, something happened in 2016 to me,
it happened to many of my colleagues and friends.
And it's happened, I think, to many of us
and to much of us as a species.
I think we're changing fast.
And I would say that when I started my practice
in 1996, right up until 2016, I was consciously looking
for primary texts as kind of teachers and mentors.
So I would look to a text by Shakespeare
or a text by Harold Pinter, or a text by TS Elliot,
or a text by Beyonce, or a text by Kanye,
or text by Imogen Heap,
or a text by any thinker, poets, author,
imaginist, visionary that I found
to fill me with possibility.
And I didn't feel able to write my own text.
I felt there was still far too much
to learn from the primary text of others.
There seemed to be so much extraordinary thought
that I was ignorant of.
So in answer to your question, most of the ideas of
the work that you may have seen all up until 2016,
all derived from a primary source.
And then I learned in 2016 when I was invited
to make a work and write the text for it,
it was a piece called Mirror Maze.
And it's gone on to lead to various other sort
of solo works in galleries,
that actually, the process when you write your own text
is not that different.
Ideas are still standing on the shoulders
of ideas that went before.
There's no such thing in my opinion,
as an idea in isolation.
I learned a lot when working
with the extraordinary theoretical physicists,
Carlo Rovelli, who in his book,
The Order of Time, reminds us that
the progress of human thought is as much influenced
by artists as it is by scientists.
And that Shar was painting
with tiny dots just at the time when the scientists were
engaged in the nature of the particle qualities
and the wave qualities of light.
And that Marcel Proust was engaged in the elasticity
of time, just when some of the theoretical physicists
were really engaged profoundly
with the nature of time.
And that, you know, there's another extraordinary example,
which was Leo Szilard, the nuclear scientist
who read HG Wells's novel in 1913, the World Set Three
and HG Wells was conjuring something entirely fictional.
He was conjuring the idea of nuclear vision.
He was conjuring the idea of a nuclear chain reaction
and this idea of an impossibly powerful force derived at
from forces within nature.
It was imaginary work, it was a work of fiction.
But Leo Zillow took it to heart
and stood at a traffic light in Southampton Row in London.
And as he watched the traffic lights change,
he had the idea that a chain reaction would be possible.
And he stopped what he was doing,
which was some work on refrigeration with Einstein.
And he dumped that and he instead devoted his life
to nuclear weapons, nuclear power.
And then he and Einstein ended up writing a letter
to Roosevelt saying, listen, if we are getting this far,
the enemy, you know, must be developing a nuclear weapon.
And that led directly to the onset of the Manhattan Project
and the development of the atomic bomb,
which then was dropped on Hiroshima and Nagasaki.
So it's just one example
of how fiction can drive our thinking forwards as fast as,
or certainly in coherence
and collaboration with science.
Fantastic, and you do a lot of design for concerts
and theater, which I imagine many people in the audience
will have seen your work as well.
How does that experience inform your approach
as an artist and a designer,
that large scale live performance
and seeing the audience respond to that?
I mean, I think the first thing to say about that
strand in my practice of designing
for large scale mass gatherings
and pop concerts is, as you know,
it's currently temporarily extinct.
So it's become even more precious to me for its,
you know, ever growing rarity each day.
It seems more precious and rare
that we ever did that, right?
The fact that we ever sat in Paris
with a hundred thousand people, you know, singing formation
album with Beyonce, that now seems kind
of otherworldly, doesn't it?
That we sang, you know, these songs together,
crunch together in these huge outdoor stadia,
it seems mythical now.
It seems like something we'll look back on
and say that's something they used to do.
It has a massive bearing on my whole practice
because there's nothing like that power
of communal energy of everybody faced in the same direction.
Everybody willing, every atom of their body
to hit the same notes.
And yet each person bringing their own very personal
connection to each particular song, you know,
each person's identification with each notes,
and each lyric comes from their own personal histories
that are unknowable to anyone else in that crowd.
So it's a once incredibly private,
personal and intimate, like no one knows the kiss you had
to that song or the wedding dance you did to that,
or your grandmother who died or whatever happened.
Nobody knows that except you.
And yet you are all doing something so, you know,
outward facing common altogether.
And it's a part of us as a species
that we're in danger of losing if we don't find a brilliant
way to keep it alive during this period
of its temporary extinction.
That's what a lot of my focus is on at the moment.
And you mentioned, obviously things have had
to change this year.
How has COVID affected your practice?
Well, that strand of the practice is not happening.
You know, there's a lot of other mercifully,
a lot of other aspects of the practice that, you know,
are getting richer and richer.
But that strand of gathering together in theater
or in a arena or in a stadium
or anywhere, any kind of mass gathering is clearly stopped.
And the big question I'm being asked by everybody, I mean,
because musicians still are making music,
they still wanna present it to their fans.
They still want to perform it.
They still want to offer a visual counterpart to it.
There's a limit to how many bedroom performances one can do.
Of course, there's the beautiful new XR technology,
but quite soon there'll be a limit to, you know,
specific ways of using that.
So it's an exciting new technique that we're,
we'll be exploring in more and more detail, I think.
But yeah, the, the practice has changed
beyond recognition in the space of seven
or eight months for sure.
Do you think there'll be sort of any long-term learnings
from this period that you might carry forward?
Oh, I think absolutely, I totally agree with that.
I think it's not just my practice
where the learnings will be, it'll be us as a species.
And my big thought about this, it's coming into sort
of shape the longer the months go on,
is I think it gives us an opportunity to analyze
what is it about what we used to do that we most miss.
Right, what was really important about it?
Why did it matter?
Why was it ever relevant to gather together?
What was it about it?
And the person, I've said this in a few talks actually,
so I, sorry to repeat myself,
but there's really,
the best answer has come from Akram Khan,
the wonderful choreographer who, when I asked him
and I said, what are you most grieving
during this period of temporary extinction?
He said, it's the quality of time passing
when I perform,
the quality of time passing when I observe
and engage in a performance.
And he reminded me that, of course, digital time
and mechanical time are incredibly recent phenomena
to our species, right?
The last kind of five minutes in terms of the existence
of the 73 millennia of our existence.
And in all pre 19th century
or even 18th century time, we would've been governed
by different types of timekeeping
and storytelling, you know, parallel engagement
in attention to narrative
and attention to music is a different way to tell the time,
you know, music literally beats out time
by a different rhythm.
It doesn't stick with the seconds
and the minutes of the clock, right?
So, that's one of the things I've singled out
as being really important
and being something that if we don't find a way
to release ourselves from digital
and mechanical means of timekeeping by these sort
of daily interactions with culture,
we may find that we do change as a species,
let alone my practice.
The actual species will evolve, you know, and change.
Because one of the other things we've been looking at a lot
is the feedback loops between the things we design
and the way that the things we design redesign us, right?
There's a wonderful book which has become very influential
to me called, Are We Human By the Design Philosophers
from MIT, Beatriz Colomina and Mark Wigley.
And they present a really compelling case
for the feedback loop that human beings design Flints,
the Flints redesign the hands of the humans,
human beings design smartphones.
The smartphones entirely redesigned
our social, cultural, physiological,
everything about us is being redesigned
by our engagement with these objects.
And that's something that, you know, we can stand by
and, you know, be complicit in and be subsumed in,
or we can really try and understand, observe
and have some agency in.
And I feel the latter is the more
appropriate response, really.
Great, it'd be interesting to see where we end up.
I think it is a huge moment of adaptation, as you say.
Yeah. Now you've got some
really exciting things coming up.
You're designing the UK Pavilion at the 2021 World Expo,
which you talked a little bit about earlier.
And you are also the artistic director
of the 2021 London Design Biennale.
What can we expect?
Well, I think the big questions we're
asking at the London Design Biennale are,
how do we as designers who make stuff,
how do we account for the stuff that we make
and how do we actually justify its existence?
You know, we have so many Design Biennale,
so many design fairs, you know,
there's a plethora of them all around the planet.
How do we really justify, even the world expo, to be honest.
You know, it's a lot of materials, it's a lot of concrete,
it's a lot of steel, it's a lot of plastic,
it's a lot of airfares.
It's a lot of stuff, right?
How do we really justify its existence?
What's it really doing?
And I guess the answer that we have to look at
is nothing should really be there
in any of these design fairs
unless it's genuinely communicating
ideas that are gonna outweigh the carbon impact
of the materials that went into producing it.
Right, so that's a piece of math you have to do.
You say, okay, well I use these materials,
but the message is so, you know, worthwhile,
it's so great in its value of how it shifts perspective
that it's worthwhile.
The other thing I think we are asking ourselves
more and more, I was on a zoom call
with all the set designers.
There's a wonderful collective they formed called
Scene Change and a group of them,
there were some from Philippines,
from Columbia, from all over the world.
The set designers are gathering on these weekly calls
and a group of them from America have begun
to suggest that maybe once set design
reemerges from hibernation, if it does,
which we hope it will.
Perhaps there'll be no longer an acceptance
of making set designs out of virgin materials.
Perhaps it'll just no longer be acceptable, you know,
in the way that smoking in pubs isn't acceptable.
Right, perhaps it'll just be only use materials
that have already had a life only use materials
that are going to have another life, you know?
And perhaps if we start with set design
and then we move on to architecture, we move on to art,
you know, into construction in general.
You know, maybe that's the kind of thing this pause
for thought could influence.
So the, yeah, the London Design Biennale
will really be looking in detail at that.
Now it's in June.
So June when we were looking last June
felt like, you know, it felt like one thing,
June, when I'm looking at it through the barrel of now,
June still feels, you know, that the presence
of the London Design Biennale will be a combination
of small numbers visiting locally, physically,
hopefully with some onsite testing
and some very safe provision.
And it will also, as you mentioned earlier,
need to have learned from this process
and it need to have a global digital reach.
And I don't think any of the practices in which I work will
reopen without that global digital reach considered.
You know, everything will open in a small,
local, physical way,
and then it'll communicate to the wider audience
through digital means.
Hopefully with more and more imaginative use
of the technology available.
Well, we look forward to joining you in whatever format
we're able to when that happens.
Thanks so much for joining us, Es.
You're so welcome, that was fun.
Thank you so much, Vicki.
Every Ancient Greek Constellation Explained
‘The Odyssey’ Cast Answer The 50 Most Searched Questions
Josh Johnson Answers The Web's Most Searched Questions
History Professor Answers: Is American Democracy Going to Die?
History Professor Answers Corruption Questions
Bernie Sanders Answers Oligarchy Questions
‘Jackass’ Cast Answer The 50 Most Searched Jackass Questions
Heike Riel: How quantum computers will drive innovation | WIRED Smarter 2019
Karyn McCluskey: How Glasgow Reduced Gang Violence Using Science
Ed Maslaveckas: Give people power over their data | WIRED Smarter 2019