Sebastian Thrun and Chris Anderson: What AI is -- and isn't
賽巴斯汀索朗及克里斯安德森: 新世代的電腦已為自己寫程式了
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,
了解一下機器學習是什麼,
and also of the concern
intelligence and machine learning
in its past until recently.
of computing and datasets
資料集規模已經達到
say, your phone,
寫程式,比如你的手機,
very long kitchen recipe,
非常非常長的廚房食譜,
turn down the temperature.
the temperature."
has 12 million lines of code.
1200 萬行的程式碼。
can cause your computer to crash.
都會造成你的電腦當機。
makes so much money.
能賺那麼多錢。
can find their own rules.
找到它們自己的規則。
deciphering, step by step,
來針對每個情況的規則
the computer examples
which recently was won by Google.
就是一個很好的例子。
you would really write down all the rules,
你得要寫下所有的規則,
residing Go champion.
the software engineer
因為軟體工程師能鬆口氣了,
where this has become really possible --
was about machine learning.
insignificant, don't read it,
were as big as a cockroach brain.
to really emulate
take advantage of the fact
much more data than people can.
進而取得優勢,
more than a million games.
一百多萬場比賽。
study a million games.
研究到一百萬場比賽。
a hundred billion web pages.
a hundred billion web pages.
一千億個網頁。
the computer can find rules
to, "If he does that, I will do that,"
「如果他那樣下,我就這樣下。」
looks like a winning pattern,
「下這裡像是獲勝的模式,
a winning pattern."
how you raise children.
你如何養育你的孩子。
對每種狀況給孩子一條規則,
giving kids a rule for every contingency
他們就會做出這個大程式。
and they have this big program.
他們會被賞巴掌或打屁股,
they get slapped or spanked,
a good grade in school,
在學校有好成績,
so much easier all of a sudden.
We just give them lots of data.
只要給它們大量資料即可。
to the spectacular improvement
into a spin-off called Voyage.
called deep learning
from Mountain View, California,
and 133 traffic lights.
途中經過 133 個交通燈號。
Google 自動駕駛汽車團隊,
the Google self-driving car team.
the world's best software engineers
最好的軟體工程師,
into the computer brain,
that often surpasses human agility.
通常都能勝過人類的機敏。
about 33 miles, an hour and a half.
大約 33 英哩,一小時半。
of this program on the left,
the computer sees as trucks and cars
所看到的卡車與汽車,
image, which is the main input here,
也就是主要的輸入,
other cars, traffic lights.
to do distance estimation.
in these kind of systems.
and so on depicted by the laser.
如樹木等等。
is centering on the camera image now.
都以攝影機影像為中心。
像是雷達和雷射,
sensors like radars and lasers
on the left thing, what is that?
for your adaptive cruise control,
供自動調整航程控制用,
how to regulate velocity
the cars in front of you are.
got an example, I think,
learning part takes place.
邊看那畫面,邊談這個。
a challenge to Udacity students
Udacity 學生的一個例子,
a self-driving car Nanodegree.
how to steer this car?"
要如何駕駛這台車?」
to get the steering right.
"It's a deep learning competition,
「這是場深度學習競賽,
like Google or Facebook,
如 Google 或臉書,
at least six months of work.
至少六個月的功夫。
100 submissions from students,
約一百件學生提交的結果,
drive on this imagery,
這影像上駕駛得還要好,
to a computer now,
to comprehend the data,
of powerful applications
強大應用程式的發展,
the other day about cancer.
CA: This is cool.
克:這很酷。
into what's happening
400,000 dollars a year,
to be a good dermatologist.
成為好的皮膚科醫生。
the machine learning version of it.
for these machine learning algorithms.
演算法的專有名詞,
by a Facebook Fellow called Yann LeCun,
研究專員揚勒丘恩所發明,
as the human brain.
but it emulates the same thing.
the visual input and extracts edges
more complicated edges
really complicated concepts.
cat faces and dog faces
at Stanford has shown is that
of skin conditions,
皮膚症狀的影像來訓練它,
that this is the case,
that we presented to our network
拿給我們的網路看,
Stanford-level dermatologists,
史丹佛水準的皮膚科醫生看,
the performance classification accuracy
網路的表現都和人類皮膚科醫生
That's a moving piece.
是個進行中的故事。
in "Nature" earlier this year
也刊在「Nature」期刊中,
dermatologists images
we had the correct classification.
by one of our collaborators.
由我們的合作者之一做的。
one of the three best, apparently,
很顯然是最好的三位之一,
"This is not skin cancer."
「這不是皮膚癌。」
a second moment, where he said,
and ran our piece of software,
執行我們的軟體,
the iPhone a little bit more than myself,"
iPhone 比相信我自己多一點點。」
to get it biopsied.
that we actually found,
would have gone unclassified,
for an app like this right now,
有很迫切的需求,
making an app that allows self-checking?
做個自我檢測的應用程式?
應用程式的信件給淹滿了,
about cancer apps,
10, 15, 20 melanomas removed,
10、15、20 個黑色素瘤,
might be overlooked, like this one,
these days, I guess.
and impress a TED audience.
來讓 TED 觀眾印象深刻。
something out that's ethical.
the assistance of a doctor
and our data holds up,
我們的資料站得住腳,
to take this kind of technology
從來沒有去過的地方。
doctors never, ever set foot.
with this army of Udacity students,
Udacity 學生大軍合作,
a different form of machine learning
一種不同形式的機器學習,
with a form of crowd wisdom.
與一種群眾智慧結合。
that could actually outperform
有時你認為那能夠勝過
even a vast company?
讓我很興奮的例子,
instances that blow my mind,
is these competitions that we run.
四十八小時就打開來用;
a self-driving car
to San Francisco on surface streets.
after seven years of Google work,
投入七年心血的成果,
and three months to do this.
用了三個月就做到這樣,
an army of students
who use crowdsourcing.
where people do bug-finding crowdsourcing
群眾外包除錯工作
in crowdsourcing.
this car in three months,
who are never hired,
and I don't even know.
maybe 9,000 answers.
which is maybe not the best thing to do.
這不見得是最好的做法。
of their education, too, which is nice.
教育的一部份,這樣想很好。
to produce amazing deep learning results.
了不起的深度學習結果。
and great machine learning is amazing.
偉大的機器學習是很驚人的。
在(TED 2017)第一天說,
the first day [of TED2017]
turned out to be two amateur chess players
是兩位業餘的棋手,
mediocre-to-good, computer programs,
with one great chess player,
和一個很棒的棋手,
you're talking about a much richer version
the fantastic panels yesterday morning,
that we sometimes confuse
是有時我們混淆了
with this kind of overlord threat,
和機器超載的威脅,
consciousness, right?
is for my AI to have consciousness.
with the dishwasher
and I don't want them.
我也不想要它們。
an augmentation of people.
of human smarts and machine smarts
結合機器的聰明,
is as old as machines are.
就像機器本身一樣老。
place because it made steam engines
做出了蒸汽引擎以及耕作設備,
that couldn't farm by itself,
it made us stronger.
will make us much, much stronger
of this for some people,
對一些人來說很駭人的部份,
scary for people is when you have
rewrite its own code,
multiple copies of itself,
if a goal is achieved and improved.
目標是否有達成或改善。
on an intelligence test.
智力測驗中得到更好的成績。
that's moderately good at that,
some sort of runaway effect
on Thursday evening,
on Friday morning,
of computers and so forth,
what I heard you say.
we had exactly this thing:
the game against itself
is a rewriting of the rules.
absolutely no concern
these are all very single-domain things.
都還是非常單一領域的東西。
that seemed nearly capable
and understand in the sense that we can,
我們的方式去閱讀及了解,
patterns of meaning.
as this broadens out,
kind of runaway effect?
I draw the line, honestly.
I don't want to downplay it --
the thing that's on my mind these days,
現在我腦中不會去想這個,
is something else.
to the present date
is because of massive numbers of Go plays,
是因為下過大量的圍棋棋譜,
or fly a plane.
or the Udacity self-driving car
Udacity 的自動駕駛汽車
and it can't do anything else.
甚至無法開摩托車。
domain-specific function,
on this thing called "general AI,"
"Hey, invent for me special relativity
狹義相對論或弦理論」的那種
and I want to acknowledge them.
"What if we can take anything repetitive
100 times as efficient?"
we all worked in agriculture
在辦公室工作,
doing repetitive things,
已經開始做重覆性工作,
of being able to take an AI,
能夠採用 AI 的邊緣,
as effective in these repetitive things.
重複性工作的效率十或五十倍。
a little terrifying to some people,
那樣的過程似乎有點嚇人,
can do this repetitive thing
is the thing that's talked about
glorious aspects of what's possible.
可能達成的輝煌面。
and it's a big issue,
by several guest speakers.
optimistic person,
back 300 years ago.
of continuous war,
or software engineer or TV anchor.
軟體工程師、電視台主播,
with a little steam engine in his pocket,
口袋中有個小蒸氣引擎,
as strong, so you can do something else."
這樣你們就可以做其它事了。」
there was no real stage,
with the cows in the stable,
和乳牛在一起,
concerned about it,
and what if the machine does this for me?"
如果讓機器來為我擠,會如何?
past progress and the benefit of it,
過去的進展和它帶來的益處,
or electricity or medical supply.
飛機、電力、醫材。
which was impossible 300 years ago.
這在三百年前是不可能的。
the same rules to the future.
同樣的規則看未來。
of my work is repetitive,
是重覆性的,
on stupid, repetitive email.
愚蠢、重覆性的電子郵件上。
that helps me get rid of this.
are insanely creative;
都極度有創意;
more than anybody else.
TED 社區更是如此。
I think you can go to your hotel maid
你可以去找你的飯店服務員,
you find a creative idea.
is to turn this creativity into action.
將創意轉化為行動。
build Google in a day?
建造出 Google,會如何?
and invent the next Snapchat,
下一個 Snapchat,會如何?
工廠工作的奴役,
in my opinion.
great side effects.
and education and shelter
醫材、教育、庇護所
affordable to all of us,
that this time it's different
這次會有所不同,
that we've used in the past
is that, not completely,
different from the kind of creativity
belief as an AI person --
any real progress on creativity
really important for people to realize,
人們很需要了解這一點,
intelligence" is so threatening,
深具威脅性的,
tossing a movie in,
the computer is our overlord,
do repetitive things.
entirely on the repetitive end.
we've become superhuman.
the Atlantic in 11 hours.
shouting back to us.
We're breaking the rules of physics.
我們打破了物理的規則。
we're going to remember everything
in my early stages of Alzheimer's.
超過 1,000 的智商。
an IQ of 1,000 or more.
spelling classes for our kids,
is that we can be super creative.
我們會超級有創意。
it's going to be painful,
of more than those jobs.
to just a new level of empowerment
60-100,000 years old, give or take --
in terms of invention,
it's a little bit older.
那就古老一些。
manufacturing, penicillin --
has gone up, not gone down, in my opinion.
已經變快了,不是變慢。
things have been invented yet. Right?
有趣的東西出來。對吧?
Hopefully, I'll change this.
希望我能改變這一點。
people laughed about. (Laughs)
Working secretly on flying cars.
暗地裡致力發明飛天車。
implant in our brain
once you have it, you'll love it.
就會愛上它。
發明的東西還沒被發明出來。
we haven't invented yet
用光束傳送到另一地。
from one location to another.
that flight wouldn't exist,
than you could run,
比你的跑步速度快,
that you can't beam a person
我們不能把一個人用光束
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