Why I draw with robots | Sougwen Chung

Translator: Ivana Korom Reviewer: Camille Martínez

翻译人员: Yanyan Hong 校对人员: Cissy Yun

Many of us here use technology in our day-to-day.

在座的各位在日常生活中 都会使用科技，

And some of us rely on technology to do our jobs.

许多人依赖科技来 进行他们的工作。

For a while, I thought of machines and the technologies that drive them

有一段时间，我认为机器和科技

as perfect tools that could make my work more efficient and more productive.

只是让我的工作更高效、高产的 完美工具。

But with the rise of automation across so many different industries,

但随着自动化技术 在各行各业的崛起，

it led me to wonder:

让我不禁试想：

If machines are starting to be able to do the work

如果机器能够完成

traditionally done by humans,

原本由人类做的工作，

what will become of the human hand?

那我们人类之手又能做些什么呢？

How does our desire for perfection, precision and automation

对完美，精确和自动化的追求

affect our ability to be creative?

是如何影响我的创造力？

In my work as an artist and researcher, I explore AI and robotics

作为艺术家和研究者， 我研究人工智能和机器人，

to develop new processes for human creativity.

以此来开发人类创造的新途径。

For the past few years,

过去几年里，

I've made work alongside machines, data and emerging technologies.

我运用机器，数据 和新兴科技进行创作。

It's part of a lifelong fascination

其中一部分永恒的魅力

about the dynamics of individuals and systems

在于人与技术间奇妙的动态，

and all the messiness that that entails.

还有其中不可避免的混乱。

It's how I'm exploring questions about where AI ends and we begin

我借此来探索人工智能与我们的界限，

and where I'm developing processes

以及探索发展

that investigate potential sensory mixes of the future.

未来感官融合的可能。

I think it's where philosophy and technology intersect.

我想这是哲学与技术的交汇点。

Doing this work has taught me a few things.

这项工作教会了我一些事。

It's taught me how embracing imperfection

它教会我拥抱不完美

can actually teach us something about ourselves.

可以帮助我们认识自我。

It's taught me that exploring art

它教会我探索艺术

can actually help shape the technology that shapes us.

能够更好的构建科技， 从而塑造自我。

And it's taught me that combining AI and robotics

它教会我将人工智能和机器人

with traditional forms of creativity -- visual arts in my case --

结合到传统的创作中—— 以我创作的视觉艺术为例——

can help us think a little bit more deeply

能够帮助我们更深入理解

about what is human and what is the machine.

何为人类，何为机器。

And it's led me to the realization

它让我意识到

that collaboration is the key to creating the space for both

在我们进步的路上，

合作是创造人与机器共存空间的关键。

as we move forward.

It all started with a simple experiment with machines,

这一切都始于 一个简单的机器实验，

called "Drawing Operations Unit: Generation 1."

实验机器叫“绘图机器：初代” （Drawing Operations Unit: Generation 1）。

I call the machine "D.O.U.G." for short.

我把它简称为道格（D.O.U.G.），

Before I built D.O.U.G,

在我建造道格之前，

I didn't know anything about building robots.

我对造机器人一无所知，

I took some open-source robotic arm designs,

我参考了一些开源的机器臂设计，

I hacked together a system where the robot would match my gestures

编成了一个系统来实现匹配手势，

and follow [them] in real time.

并实时模仿它们。

The premise was simple:

前提很简单：

I would lead, and it would follow.

我画，而它会学我。

I would draw a line, and it would mimic my line.

我画一条线， 它也会跟着我画一条线。

So back in 2015, there we were, drawing for the first time,

回到 2015 年，那是我们第一次

in front of a small audience in New York City.

在纽约的一小群观众前作画。

The process was pretty sparse --

整个过程非常冷清——

no lights, no sounds, nothing to hide behind.

没有灯光，没有音效， 也没有什么悬念。

Just my palms sweating and the robot's new servos heating up.

只有手掌冒出的冷汗 和机器臂不断升高的温度。

(Laughs) Clearly, we were not built for this.

（笑声）显然， 这不是我们想要的效果。

But something interesting happened, something I didn't anticipate.

但有趣的事发生了， 完全出乎意料。

See, D.O.U.G., in its primitive form, wasn't tracking my line perfectly.

初代的道格并没有 完美地模仿我画的线条，

While in the simulation that happened onscreen

在计算器模拟中显示

it was pixel-perfect,

它的模仿事精确完美的，

in physical reality, it was a different story.

但到了现实中，却并非如此。

It would slip and slide and punctuate and falter,

它会滑动，会卡顿，会晃动，

and I would be forced to respond.

于是我不得不附和它的线条。

There was nothing pristine about it.

它的状态不完美，

And yet, somehow, the mistakes made the work more interesting.

而这些失误让作品更加有趣，

The machine was interpreting my line but not perfectly.

机器在模仿我的线条， 但是并不完美，

And I was forced to respond.

于是变成我在附和机器。

We were adapting to each other in real time.

我们不断地实时熟悉彼此。

And seeing this taught me a few things.

看到这些，教会了我一些事，

It showed me that our mistakes actually made the work more interesting.

我们的失误，实际上 让我们的作品更加有趣，

And I realized that, you know, through the imperfection of the machine,

我从机器的不完美中意识到，

our imperfections became what was beautiful about the interaction.

我们的不完美成就了这互动之美。

And I was excited, because it led me to the realization

而我很兴奋，因为它让我意识到

that maybe part of the beauty of human and machine systems

或许人类和机器系统的美妙之一

is their shared inherent fallibility.

正是他们共同的、固有的不完美。

For the second generation of D.O.U.G.,

对于第二代的道格，

I knew I wanted to explore this idea.

我知道我要探索这个想法，

But instead of an accident produced by pushing a robotic arm to its limits,

我并非打算通过放大机器臂的失误，

I wanted to design a system that would respond to my drawings

而是想要设计一个系统 能够以出其不意的方式

in ways that I didn't expect.

回应我的画作。

So, I used a visual algorithm to extract visual information

所以，我运用一个视觉算法 来提取我几十年来的

from decades of my digital and analog drawings.

数字和实体绘图中的视觉样本信息，

I trained a neural net on these drawings

以此我训练了一个神经网络

in order to generate recurring patterns in the work

优化机器的循环模式，

that were then fed through custom software back into the machine.

视觉样本由经专门的 软件处理导入机器。

I painstakingly collected as many of my drawings as I could find --

于是我煞费苦心地 收集我的所有的画作——

finished works, unfinished experiments and random sketches --

成品，半成品，随手简笔画——

and tagged them for the AI system.

把它们标记给人工智能系统。

And since I'm an artist, I've been making work for over 20 years.

作为一位艺术家， 我作画超过了 20 年，

Collecting that many drawings took months,

所以收集这些画作花了好多个月，

it was a whole thing.

这是个大工程。

And here's the thing about training AI systems:

说到训练人工智能：

it's actually a lot of hard work.

这其实大费功夫。

A lot of work goes on behind the scenes.

幕后的工作很多很多，

But in doing the work, I realized a little bit more

但在其中，我对人工智能的构造

about how the architecture of an AI is constructed.

更深入了解了一些。

And I realized it's not just made of models and classifiers

我意识到它不仅是 神经网络的

for the neural network.

模型和分屏器。

But it's a fundamentally malleable and shapable system,

它是一个可延展的、可塑的系统，

one in which the human hand is always present.

人类的手始终参与其中。

It's far from the omnipotent AI we've been told to believe in.

它不再是我们认为的 无所不能的人工智能。

So I collected these drawings for the neural net.

所以，我收集画作以训练神经网络，

And we realized something that wasn't previously possible.

而且我们意识到 前所未有的事情发生了，

My robot D.O.U.G. became a real-time interactive reflection

我对机器人道格 在实时交互创作中，

of the work I'd done through the course of my life.

对我过去人生几十年的作品做出回应。

The data was personal, but the results were powerful.

数据源于我个人， 但结果却很有力。

And I got really excited,

我感到非常兴奋，

because I started thinking maybe machines don't need to be just tools,

因为我开始想或许机器不该只是工具，

but they can function as nonhuman collaborators.

它还可以是非人的合作者。

And even more than that,

再进一步想，

I thought maybe the future of human creativity

也许未来的人类创作

isn't in what it makes

不在于作品本身，

but how it comes together to explore new ways of making.

而在于对艺术诞生新方式的探索。

So if D.O.U.G._1 was the muscle,

所以，如果道格初代是肌肉，

and D.O.U.G._2 was the brain,

那么道格二代就是大脑，

then I like to think of D.O.U.G._3 as the family.

然后我想道格三代就是家人。

I knew I wanted to explore this idea of human-nonhuman collaboration at scale.

我知道我想要将对 人类和非人类合作的想法放大。

So over the past few months,

于是再过去的几个月里，

I worked with my team to develop 20 custom robots

我和团队造出了 20 个定制的机器人

that could work with me as a collective.

与我集体创作。

They would work as a group,

它们像团队一样工作，

and together, we would collaborate with all of New York City.

我们共同与整个纽约市携手合作，

I was really inspired by Stanford researcher Fei-Fei Li,

斯坦福大学的研究员李飞飞 激发了我对灵感，

who said, "if we want to teach machines how to think,

她说，"若像教机器如何思考，

we need to first teach them how to see."

先要教它们如何看见。"

It made me think of the past decade of my life in New York,

这让我想起了过去 十年的纽约生活，

and how I'd been all watched over by these surveillance cameras around the city.

城市上空的监控摄像头监视着我，

And I thought it would be really interesting

如果我用它们来训练

if I could use them to teach my robots to see.

我的机器人的视觉， 那会非常有趣。

So with this project,

所以在这个项目中，

I thought about the gaze of the machine,

我思考机器对我们的凝视，

and I began to think about vision as multidimensional,

于是我开始将视觉看成多元化的，

as views from somewhere.

视作来自某处的视点。

We collected video

我们收集视频，

from publicly available camera feeds on the internet

从网络上公共摄像头的影片

of people walking on the sidewalks,

到行人在路上走的片段，

cars and taxis on the road,

道路上的汽车和出租，

all kinds of urban movement.

城市中各种车水马龙的片段。

We trained a vision algorithm on those feeds

基于一种“光流技术”，

based on a technique called "optical flow,"

我们训练了一种视觉算法，

to analyze the collective density,

来分析收集到的人流密度，

direction, dwell and velocity states of urban movement.

城市流动的方向， 速度状态以及居住方式。

Our system extracted those states from the feeds as positional data

我们的系统从海量的 位置数据中提取这些信息，

and became pads for my robotic units to draw on.

我们的机器人依靠这些信息来作画，

Instead of a collaboration of one-to-one,

与之前的一对一合作不同，

we made a collaboration of many-to-many.

我们实现了多对多的合作。

By combining the vision of human and machine in the city,

通过结合城市中人类与机器的视角，

we reimagined what a landscape painting could be.

我们重构了一个景观绘图可能的样子。

Throughout all of my experiments with D.O.U.G.,

在我和道格所有的实验中，

no two performances have ever been the same.

没有哪两次的呈现是相同的，

And through collaboration,

而且通过合作，

we create something that neither of us could have done alone:

我们创作了我们 无法独自实现的事情，

we explore the boundaries of our creativity,

我们共同探索了创造力的边界，

human and nonhuman working in parallel.

人类和非人类并肩工作。

I think this is just the beginning.

我想这才是开始，

This year, I've launched Scilicet,

今年，我创办了 Scilicet，

my new lab exploring human and interhuman collaboration.

这个新实验室旨在探索 人类和非人类间的合作，

We're really interested in the feedback loop

我们对个体，人工和生态系统

between individual, artificial and ecological systems.

之间的反馈关系非常感兴趣。

We're connecting human and machine output

我们将人类和机器与

to biometrics and other kinds of environmental data.

生物特征识别和其他环境数据相结合。

We're inviting anyone who's interested in the future of work, systems

我们邀请任何对未来的作品、系统

and interhuman collaboration

和人际间合作感兴趣的人

to explore with us.

和我们共同探索。

We know it's not just technologists that have to do this work

我们知道不仅是科技工作者肩负使命，

and that we all have a role to play.

所有人都可以参与其中。

We believe that by teaching machines

我们坚信通过教授机器

how to do the work traditionally done by humans,

如何去完成人类的传统工作，

we can explore and evolve our criteria

我们就能不断探索和创新

of what's made possible by the human hand.

超越人类之手所能达到的可能。

And part of that journey is embracing the imperfections

这段旅程之一便是拥抱不完美，

and recognizing the fallibility of both human and machine,

发现人类和机器共有的缺憾，

in order to expand the potential of both.

才能更好的拓展我们共同的潜能。

Today, I'm still in pursuit of finding the beauty

今天，我仍在追寻人类和

in human and nonhuman creativity.

非人类协作的美妙之处。

In the future, I have no idea what that will look like,

在未来，我不知道会怎样，

but I'm pretty curious to find out.

但是我满怀好奇去寻找答案。

Thank you.

谢谢。

(Applause)

（掌声）