Sebastian Thrun and Chris Anderson: What AI is -- and isn't
세바스찬 쓰런 (Sebastian Thrun), 크리스 앤더슨 (Chris Anderson): 새로운 세대의 컴퓨터는 스스로 프로그래밍한다
Sebastian Thrun is a passionate technologist who is constantly looking for new opportunities to make the world better for all of us. Full bioChris Anderson - TED Curator
After a long career in journalism and publishing, Chris Anderson became the curator of the TED Conference in 2002 and has developed it as a platform for identifying and disseminating ideas worth spreading. Full bio
Double-click the English transcript below to play the video.
what machine learning is,
데 도움을 주실 분을 모셨습니다.
and also of the concern
intelligence and machine learning
in its past until recently.
of computing and datasets
이루어졌기 때문이죠.
say, your phone,
만든다고 가정해봅시다.
very long kitchen recipe,
조리법을 적게 합니다.
turn down the temperature.
온도를 내려라.
the temperature."
하는 식으로 말이죠.
has 12 million lines of code.
소스 코드로 되어 있습니다.
줄 정도로 구성됩니다.
can cause your computer to crash.
컴퓨터를 고장나게 할 수 있죠.
makes so much money.
돈을 많이 벌 수 밖에 없죠.
can find their own rules.
스스로 규칙을 찾아냅니다.
deciphering, step by step,
해독해야 했습니다.
the computer examples
예시만 던져주고
which recently was won by Google.
좋은 예시가 되겠네요.
you would really write down all the rules,
모든 규칙을 적어주어야 하겠죠.
넘는 기보를 공부해서
residing Go champion.
바둑 기사를 이겼죠.
the software engineer
개발하는 사람이
한다는 부담을 덜어줍니다.
where this has become really possible --
가능하게 된 계기는 --
was about machine learning.
기계학습에 관해 썼습니다.
insignificant, don't read it,
읽지는 말아주세요.
were as big as a cockroach brain.
to really emulate
사람과 거의 비슷하게
take advantage of the fact
더 많은 데이터를
much more data than people can.
우위에 있습니다.
more than a million games.
기보를 학습할 수 있었던 거죠.
study a million games.
사람에겐 불가능한 일이지요.
a hundred billion web pages.
웹 페이지를 살펴 봅니다.
a hundred billion web pages.
the computer can find rules
찾을 수 없던 규칙들을
to, "If he does that, I will do that,"
나는 저렇게 해야지" 라고
looks like a winning pattern,
"이런 식으로 하면 이기는구나."
a winning pattern."
라고 생각하는 거군요.
how you raise children.
아이를 키울 때를 생각해보세요.
giving kids a rule for every contingency
일일이 지시를 내리지는 않죠.
and they have this big program.
큰 프로그램을 갖게 되죠.
they get slapped or spanked,
다시 일어서고, 또 혼나기도 합니다.
a good grade in school,
긍정적인 경험도 할 겁니다.
so much easier all of a sudden.
갑자기 너무 쉬워졌죠.
We just give them lots of data.
정보를 넘겨주면 됩니다.
to the spectacular improvement
혁신적으로 발전하게 된 데에
into a spin-off called Voyage.
새 버전을 만들었죠.
called deep learning
훈련시켰습니다.
from Mountain View, California,
따라 주행하며
and 133 traffic lights.
133개의 신호등을 지나쳤죠.
the Google self-driving car team.
자율주행 자동차 팀에서 일했을 때
the world's best software engineers
그냥 훈련을 받았습니다.
into the computer brain,
that often surpasses human agility.
운전하기도 하는 겁니다.
about 33 miles, an hour and a half.
만에 사람의 도움없이 주행했습니다.
of this program on the left,
왼쪽에는 컴퓨터가 지각하는 모습이죠?
the computer sees as trucks and cars
image, which is the main input here,
영상이 컴퓨터의 메인 입력이죠.
other cars, traffic lights.
신호등을 인식합니다.
to do distance estimation.
레이더도 달려있습니다.
in these kind of systems.
흔히 쓰이는 장비입니다.
도표입니다.
and so on depicted by the laser.
그린 것을 보실 수 있습니다.
is centering on the camera image now.
카메라 영상 쪽 입니다.
sensors like radars and lasers
바꾸는 중입니다.
on the left thing, what is that?
초록점은 뭔가요?
for your adaptive cruise control,
속도를 조절하는 적응형 순항 장치가
how to regulate velocity
있는지 표시한 겁니다.
the cars in front of you are.
이해하도록 말이죠.
got an example, I think,
어떻게 학습을 하는지에 대한
learning part takes place.
a challenge to Udacity students
자동차 '나노디그리'라고 부르는
a self-driving car Nanodegree.
how to steer this car?"
알아낼 수 있겠니?" 라고 물었습니다.
to get the steering right.
방향을 맞추기 힘듭니다.
"It's a deep learning competition,
대회이고, AI 대회야."
like Google or Facebook,
기업에서 같은 작업을 했다면
at least six months of work.
100 submissions from students,
답안을 제출하였습니다.
답안은 거의 완벽에 가까웠습니다.
drive on this imagery,
같은 방법론입니다.
to a computer now,
to comprehend the data,
of powerful applications
많이 개발되는 데에
the other day about cancer.
게 있다고 하셨죠.
CA: This is cool.
크리스: 대단한 영상이에요.
into what's happening
어떻게 인공지능이 활용 되는지
400,000 dollars a year,
to be a good dermatologist.
되려면 10년이 넘게 걸립니다.
the machine learning version of it.
머신러닝의 결과물 입니다.
for these machine learning algorithms.
알고리즘을 부르는 말입니다.
by a Facebook Fellow called Yann LeCun,
이라는 페이스북 개발자의 작품입니다.
as the human brain.
but it emulates the same thing.
비슷하게 모방한 것이죠.
the visual input and extracts edges
입력받은 후 그 안에서
more complicated edges
really complicated concepts.
개념들을 다룰 수 있어요.
cat faces and dog faces
고양이와 개의 얼굴을
at Stanford has shown is that
데리고있는 학생들은
of skin conditions,
9천 장의 피부 사진을 가지고
피부과 전문의 만큼의
것을 보여줬습니다.
that this is the case,
있다는 것을 증명하기 위해
that we presented to our network
독립적인 데이터셋과
Stanford-level dermatologists,
피부과 전문의들의 데이타셋을 모아서
the performance classification accuracy
결과와 비슷하거나 더 뛰어난
That's a moving piece.
in "Nature" earlier this year
지에도 실린 내용인데요.
dermatologists images
측정해보았죠.
분류가 끝난 것들입니다.
we had the correct classification.
생체 검사가 끝난 것들만 모았죠.
by one of our collaborators.
스탠포드에서 찍은 사진이에요.
피부과 의사인데요.
one of the three best, apparently,
세 손가락 중에 하나에 드는 사람이죠.
"This is not skin cancer."
"피부암이 아니다"고 했어요.
a second moment, where he said,
and ran our piece of software,
우리가 만든 소프트웨어
암이라고 진단했어요.
the iPhone a little bit more than myself,"
to get it biopsied.
사진을 연구소로 보냈습니다.
that we actually found,
would have gone unclassified,
for an app like this right now,
어마어마한 숫자의 사람들이
이용하고 싶어할 것 같은데요.
making an app that allows self-checking?
만들 계획을 갖고 있나요?
about cancer apps,
아픈 사연을 가진 사람들이 보낸
10, 15, 20 melanomas removed,
흑색종을 제거하고도
might be overlooked, like this one,
두려워합니다.
these days, I guess.
and impress a TED audience.
깊은 인상을 심어주는 일은 쉽지만
something out that's ethical.
섣불리 움직일 수 없습니다.
the assistance of a doctor
and our data holds up,
to take this kind of technology
doctors never, ever set foot.
with this army of Udacity students,
함께 작업하기 때문에
a different form of machine learning
머신 러닝 기술과 집단 지성을
with a form of crowd wisdom.
한다는 말인가요?
that could actually outperform
능력을 앞설 수 있다고 보십니까?
even a vast company?
instances that blow my mind,
순간들이 있죠.
is these competitions that we run.
진행하는 경연대회를 말하는 거에요.
a self-driving car
자동차를 만들 수 있습니다.
to San Francisco on surface streets.
샌프란시스코까지 주행 가능합니다.
after seven years of Google work,
비교할 바는 아니지만
and three months to do this.
고작 세 달만에 완성했죠.
an army of students
who use crowdsourcing.
우리뿐만이 아닙니다.
크라우드소싱을 활용하죠.
호텔에 적용했습니다.
where people do bug-finding crowdsourcing
활용하는 예는 얼마든지 있습니다.
in crowdsourcing.
this car in three months,
who are never hired,
해고하지도 않는 직원이
and I don't even know.
maybe 9,000 answers.
한다는 규정도 없습니다.
댓가를 지불하죠.
which is maybe not the best thing to do.
행동하고 있습니다. 최선은 아니죠.
of their education, too, which is nice.
경험이 될 거라고 생각합니다.
to produce amazing deep learning results.
놀라운 결과물을 만들 수 있죠.
and great machine learning is amazing.
만난 결과는 정말 놀랍습니다.
the first day [of TED2017]
TED2017의 첫째날 한 말인데요.
turned out to be two amateur chess players
두 명의 아마추어 체스 기사
mediocre-to-good, computer programs,
컴퓨터 프로그램을 이용하여
with one great chess player,
뛰어난 체스 기사를 이겼다는 거요.
you're talking about a much richer version
the fantastic panels yesterday morning,
들으셨던 것처럼
사람들의 반응 등
that we sometimes confuse
우려를 표하는 부분은
with this kind of overlord threat,
갖는 자의식이 로봇의 인간 지배라는
consciousness, right?
수준까지 도달했냐하는 거죠.
is for my AI to have consciousness.
의식을 심어주는 것입니다.
with the dishwasher
말하는 상황을 원하지 않아요.
and I don't want them.
않고 원하지도 않아요.
an augmentation of people.
돕기 위한 것이었습니다.
정확했다고 생각해요.
of human smarts and machine smarts
is as old as machines are.
인류를 위한 것이었습니다.
place because it made steam engines
that couldn't farm by itself,
이것들은 인간을 대체하지 않았고
it made us stronger.
좀 더 용이하게 해주었습니다.
will make us much, much stronger
바람도 인류를 더욱
좀 더 이야기할 것이 있는데요.
of this for some people,
분들을 위해서 말이죠.
scary for people is when you have
있는 컴퓨터를 만들고
rewrite its own code,
재생산할 수 있는 능력은 물론
multiple copies of itself,
if a goal is achieved and improved.
확인하는 겁니다.
on an intelligence test.
을 내는 것이라고 치죠.
that's moderately good at that,
꽤 훌륭하잖아요.
과정을 반복하는 거죠.
some sort of runaway effect
못하는 상황과 마주치지 않을까 하는
on Thursday evening,
on Friday morning,
of computers and so forth,
일어날 가능성이죠.
what I heard you say.
we had exactly this thing:
다음 사실을 목격했습니다.
the game against itself
바둑을 둘 수 있고
is a rewriting of the rules.
규칙을 다시 사용하는 것입니다.
absolutely no concern
지배할 거라는 걱정은
these are all very single-domain things.
대해서만 이야기하고 있어요.
that seemed nearly capable
통과할 수 있을 정도의
and understand in the sense that we can,
patterns of meaning.
as this broadens out,
kind of runaway effect?
상황이 생길 수 있지 않을까요?
I draw the line, honestly.
않다고 생각합니다.
I don't want to downplay it --
그것까지 부정하고 싶진 않아요.
the thing that's on my mind these days,
지금 염두에 두고 있는 문제는 아닙니다.
is something else.
일아나고 있다고 생각해요.
to the present date
분야에 한정되어 있구요.
한다는 발상에 기대어
is because of massive numbers of Go plays,
엄청난 횟수의 대국을 두기 때문이죠.
or fly a plane.
비행기를 몰 수 없어요.
or the Udacity self-driving car
and it can't do anything else.
그 이상의 것은 불가능합니다.
domain-specific function,
활용 가능하고
마찬가지입니다.
on this thing called "general AI,"
끈 이론을 만들어 봐"라고
"Hey, invent for me special relativity
라고 하는 영역에서는
and I want to acknowledge them.
제대로 확인시켜 주고 싶기 때문입니다.
"What if we can take anything repetitive
100 times as efficient?"
일할 수 있다면 어떻게 될까?"
we all worked in agriculture
사무실에서 일하고
원숭이처럼 되버렸어요.
doing repetitive things,
of being able to take an AI,
as effective in these repetitive things.
효율을 낼 수 있다고 생각합니다.
a little terrifying to some people,
이들에게는 두려움으로 다가올 것 같아요.
can do this repetitive thing
수 있게 된다면 말이죠.
is the thing that's talked about
자율주행 자동차 등으로
않을까요?
희망찬 미래에 도달하기 전에
glorious aspects of what's possible.
and it's a big issue,
정말 큰 문제입니다.
by several guest speakers.
그 점을 지적했어요.
optimistic person,
사람임을 고백합니다.
back 300 years ago.
of continuous war,
전쟁에 시달렸고
아무것도 존재하지 않았습니다.
or software engineer or TV anchor.
TV 앵커 같은 직업이요.
with a little steam engine in his pocket,
작은 증기 기관을 꺼내 보여주며
이 증기기관을 사용하면
as strong, so you can do something else."
남은 시간을 활용할 수 있어."
there was no real stage,
with the cows in the stable,
소들을 둘러보는 동안
concerned about it,
있는 게 하나 있는데
and what if the machine does this for me?"
기계가 대신해주면 어떨까?"
past progress and the benefit of it,
좋은 점을 취할 줄 알았어요.
or electricity or medical supply.
의료기기 같은 것들 말이에요.
which was impossible 300 years ago.
300년 전에는 불가능했던 일입니다.
the same rules to the future.
똑같으리라 가정할 수는 없겠죠.
of my work is repetitive,
on stupid, repetitive email.
똑같은 이메일 작성에 허비해요.
that helps me get rid of this.
있다면 정말 좋겠다고 생각합니다.
are insanely creative;
창의적이라고 믿으니까요.
more than anybody else.
여러분은 그럴거라 생각합니다.
I think you can go to your hotel maid
호텔의 종업원과 어울리며
you find a creative idea.
창의적인 생각이 떠오르는 겁니다.
is to turn this creativity into action.
행동으로 옮길 수 있게 되는 겁니다.
build Google in a day?
만들 수 있다면 어떨까요?
and invent the next Snapchat,
'스냅챗'을 발명하고
될 것이라는 점입니다.
해방시킬 수 있었고
in my opinion.
나아질거라 생각합니다.
great side effects.
and education and shelter
주거, 운송수단등이
affordable to all of us,
that this time it's different
다르다고 했던 것 기억하시나요?
that we've used in the past
is that, not completely,
different from the kind of creativity
belief as an AI person --
하는 사람으로서 제가 가진 신념입니다.
any real progress on creativity
really important for people to realize,
위협적인 느낌과
intelligence" is so threatening,
tossing a movie in,
지배하는 등의 설정 때문에
the computer is our overlord,
단순한 '기술'입니다.
do repetitive things.
덜어주는 기술적 진보죠.
entirely on the repetitive end.
분야에서만 이뤄졌습니다.
한정되어 있기 때문에
아닐 것 같습니다.
we've become superhuman.
우리는 이미 초인간입니다.
the Atlantic in 11 hours.
가로지를 수 있습니다.
shouting back to us.
연락을 받는 것 역시 가능합니다.
We're breaking the rules of physics.
우리는 물리법칙을 무너뜨리고 있어요.
we're going to remember everything
여러분은 보고 들은
기억하게 될 것이구요.
in my early stages of Alzheimer's.
좋은 일이죠.
기억이 안 나네요.
an IQ of 1,000 or more.
1000이나 그 이상이 될 겁니다.
spelling classes for our kids,
수업을 듣지 않아도 됩니다.
생길 일이 없어질 거니까요.
is that we can be super creative.
창의적인 인간이 될 수 있습니다.
정도 발생할 것이고
it's going to be painful,
of more than those jobs.
더 많은 것을 얻게 될 것이다.
to just a new level of empowerment
발견을 제공할 수 있는
60-100,000 years old, give or take --
in terms of invention,
나온 것들입니다.
it's a little bit older.
manufacturing, penicillin --
탄생한 것들입니다.
발명을 할 수 있다는 의미로 여겨집니다.
has gone up, not gone down, in my opinion.
빨라지고 있다고 생각합니다.
things have been invented yet. Right?
1%밖에 이뤄지지 않았다고 생각해요.
Hopefully, I'll change this.
하지만 저는 계속 도전할 겁니다.
people laughed about. (Laughs)
발명의 전형적인 예죠. (웃음)
Working secretly on flying cars.
하늘을 나는 자동차를 개발한다는 거요.
두 배만큼 살지도 못합니다.
implant in our brain
바로 취득할 수 있도록
일도 아직 멀었죠.
once you have it, you'll love it.
마음에 쏙 들게 될거라는 겁니다.
we haven't invented yet
생각하는 것들 중에
from one location to another.
순간이동 하는 장치 말입니다.
that flight wouldn't exist,
존재하지 못할거라 생각했고
than you could run,
더 빠르게 움직이면
that you can't beam a person
순간이동을 할 수 없다 라는게
혜안과 재능을 공유해 주셔서
and your brilliance.
ABOUT THE SPEAKERS
Sebastian Thrun - Educator, entrepreneurSebastian Thrun is a passionate technologist who is constantly looking for new opportunities to make the world better for all of us.
Why you should listen
Sebastian Thrun is an educator, entrepreneur and troublemaker. After a long life as a professor at Stanford University, Thrun resigned from tenure to join Google. At Google, he founded Google X, home to self-driving cars and many other moonshot technologies. Thrun also founded Udacity, an online university with worldwide reach, and Kitty Hawk, a "flying car" company. He has authored 11 books, 400 papers, holds 3 doctorates and has won numerous awards.
Sebastian Thrun | Speaker | TED.com
Chris Anderson - TED Curator
After a long career in journalism and publishing, Chris Anderson became the curator of the TED Conference in 2002 and has developed it as a platform for identifying and disseminating ideas worth spreading.
Why you should listen
Chris Anderson is the Curator of TED, a nonprofit devoted to sharing valuable ideas, primarily through the medium of 'TED Talks' -- short talks that are offered free online to a global audience.
Chris was born in a remote village in Pakistan in 1957. He spent his early years in India, Pakistan and Afghanistan, where his parents worked as medical missionaries, and he attended an American school in the Himalayas for his early education. After boarding school in Bath, England, he went on to Oxford University, graduating in 1978 with a degree in philosophy, politics and economics.
Chris then trained as a journalist, working in newspapers and radio, including two years producing a world news service in the Seychelles Islands.
Back in the UK in 1984, Chris was captivated by the personal computer revolution and became an editor at one of the UK's early computer magazines. A year later he founded Future Publishing with a $25,000 bank loan. The new company initially focused on specialist computer publications but eventually expanded into other areas such as cycling, music, video games, technology and design, doubling in size every year for seven years. In 1994, Chris moved to the United States where he built Imagine Media, publisher of Business 2.0 magazine and creator of the popular video game users website IGN. Chris eventually merged Imagine and Future, taking the combined entity public in London in 1999, under the Future name. At its peak, it published 150 magazines and websites and employed 2,000 people.
This success allowed Chris to create a private nonprofit organization, the Sapling Foundation, with the hope of finding new ways to tackle tough global issues through media, technology, entrepreneurship and, most of all, ideas. In 2001, the foundation acquired the TED Conference, then an annual meeting of luminaries in the fields of Technology, Entertainment and Design held in Monterey, California, and Chris left Future to work full time on TED.
He expanded the conference's remit to cover all topics, including science, business and key global issues, while adding a Fellows program, which now has some 300 alumni, and the TED Prize, which grants its recipients "one wish to change the world." The TED stage has become a place for thinkers and doers from all fields to share their ideas and their work, capturing imaginations, sparking conversation and encouraging discovery along the way.
In 2006, TED experimented with posting some of its talks on the Internet. Their viral success encouraged Chris to begin positioning the organization as a global media initiative devoted to 'ideas worth spreading,' part of a new era of information dissemination using the power of online video. In June 2015, the organization posted its 2,000th talk online. The talks are free to view, and they have been translated into more than 100 languages with the help of volunteers from around the world. Viewership has grown to approximately one billion views per year.
Continuing a strategy of 'radical openness,' in 2009 Chris introduced the TEDx initiative, allowing free licenses to local organizers who wished to organize their own TED-like events. More than 8,000 such events have been held, generating an archive of 60,000 TEDx talks. And three years later, the TED-Ed program was launched, offering free educational videos and tools to students and teachers.
Chris Anderson | Speaker | TED.com