Jeremy Howard: The wonderful and terrifying implications of computers that can learn
Jeremy Howard: Öğrenebilen bilgisayarların müthiş ve ürkütücü uygulamaları
Jeremy Howard imagines how advanced machine learning can improve our lives. Full bio
Double-click the English transcript below to play the video.
to get a computer to do something new,
that haven't done it yourself,
her şeyi amacına
the computer to do
that you don't know how to do yourself,
to be a great challenge.
by this man, Arthur Samuel.
program tasarlayarak
how to be better than you at checkers?
olmasını sağlayabilirsin?
against itself thousands of times
and in fact, by 1962,
the Connecticut state champion.
the father of machine learning,
makine öğrenmenin
learning practitioner.
machine learning practictioners.
previously unsolved problems,
hundreds of times.
I was able to find out
can do in the past, can do today,
machine learning commercially was Google.
possible to find information
on machine learning.
commercial successes of machine learning.
olarak başarılı uygulaması oldu.
products that you might like to buy,
şirketler bazen size
who your friends might be
gösterir ve siz
the power of machine learning.
learned how to do this from data
the two world champions at "Jeopardy,"
and complex questions like this one.
from this city's national museum in 2003
(along with a lot of other stuff)"]
to see the first self-driving cars.
bu sistem ile çalışır.
the difference between, say,
well, that's pretty important.
those programs by hand,
this is now possible.
over a million miles
öğrenebildiğini
don't know how to do ourselves,
I've seen of machine learning
called Geoffrey Hinton
automatic drug discovery.
is not just that they beat
or the international academic community,
in chemistry or biology or life sciences,
çözmeleri değil bunu
called deep learning.
bir algoritma kullandılar.
the success was covered
article a few weeks later.
kapağında yer aldı.
here on the left-hand side.
inspired by how the human brain works,
çalışmasından
oluşturulmuş bir algoritmadır
teorik kısıtlamalar içermez.
on what it can do.
computation time you give it,
elde edersiniz.
showed in this article
result of deep learning
can listen and understand.
to take in this process
of information from many Chinese speakers
and converts it into Chinese language,
bir sistem oluşturduk.
an hour or so of my own voice
so that it would sound like me.
a machine learning conference in China.
at academic conferences
at TEDx conferences, feel free.
was happening with deep learning.
Teşekkürler.
was deep learning.
öğrenme ile yapıldı.
in the top right, deep learning,
metin derin öğrenmeyle,
was deep learning as well.
de derin öğrenme ile yapıldı.
this extraordinary thing.
neredeyse
can seem to do almost anything,
it had also learned to see.
görmeyi de öğrenmiş.
İşaretleri Karşılaştırması
Recognition Benchmark,
to recognize traffic signs like this one.
işaretleri algılamayı öğrendi.
recognize the traffic signs
sıralamasına bakıldığında
it was better than people,
better than people.
they had a deep learning algorithm
videolarını izleyen
veriyi sıkıştırabilen
on 16,000 computers for a month,
olduğunu açıkladı.
about concepts such as people and cats
ayrı ayrı insan ve kedi
that humans learn.
by being told what they see,
anlatılmasıyla değil
what these things are.
who we saw earlier,
gördüğümüz Geoffrey Hinton
bakarak neyin fotoğrafları
from one and a half million images
to a six percent error rate
an extraordinarily good job of this,
harika işler yapıyor
location in France in two hours,
that they fed street view images
to recognize and read street numbers.
algoritmasında kullandılar.
it would have taken before:
the Chinese Google, I guess,
to Baidu's deep learning system,
has understood what that picture is
bulduğunu görebilirsiniz.
have similar backgrounds,
benzer arka planlara,
at the text of a web page.
metnine bakmak gibi değil.
really understand what they see
of images in real time.
now that computers can see?
ne anlama geliyor?
that computers can see.
bundan daha fazlasını yaptı.
has done more than that.
with deep learning algorithms.
algoritmalarıyla anlaşılabilir.
showing the red dot at the top
is expressing negative sentiment.
his ifade ettiğini anlıyor.
is near human performance
and what it is saying about those things.
been used to read Chinese,
okumada da kullanıldı.
Chinese speaker level.
out of Switzerland
or understand any Chinese.
tarafından İsviçre'de geliştirildi.
öğrenme kullanma bu işte
in the world for this,
human understanding.
put together at my company
all this stuff together.
toplamayı gösteriyor.
have no text attached,
these pictures
to the text that I'm writing.
understanding my sentences
cümlelerimi anlama
something like this on Google,
gördüğünüzü düşünüyorum;
and it will show you pictures,
resimler gösterir ama
searching the webpage for the text.
yazdıklarınızı aramaktır.
understanding the images.
oldukça farklı.
have only been able to do
can not only see but they can also read,
can understand what they hear.
I'm going to tell you they can write.
using a deep learning algorithm yesterday.
öğrenme algoritmasıyla
out of Stanford generated.
geliştirilen metinler var.
resimleri tanımlamak
to describe each of those pictures.
algoritmasıyla geliştirildi .
a man in a black shirt playing a guitar.
it's seen black before,
this novel description of this picture.
performance here, but we're close.
the computer-generated caption
insanlar, bilgisayar
well past human performance
to very exciting opportunities.
Boston'da bir takım,
that they had discovered
olarak ilişkili tümör
make a prognosis of a cancer.
looking at tissues under magnification,
oranlarını ölçmede
a machine learning-based system
than human pathologists
bir sistem geliştirdi.
for cancer sufferers.
were the predictions more accurate,
bir bilim geliştirdiler.
that humans can understand.
kliniksel indikatörlerdi.
that the cells around the cancer
etrafındaki hücrelerin kanser
the cancer cells themselves
had been taught for decades.
they were systems developed
and machine learning experts,
we're now beyond that too.
identifying cancerous areas
tanımlamak üzerine.
can identify those areas more accurately,
as human pathologists,
using no medical expertise
no background in the field.
about as accurately as humans can,
with deep learning
background in medicine.
no previous background in medicine,
to start a new medical company,
biraz korkmuştum ama
that it ought to be possible
tekniklerini kullanarak
using just these data analytic techniques.
has been fantastic,
but from the medical community,
orta bölümünü alıp
the middle part of the medical process
as much as possible,
alana yoğunlaştırabiliriz.
what they're best at.
to generate a new medical diagnostic test
three minutes by cutting some pieces out.
creating a medical diagnostic test,
göstermek yerine
a diagnostic test of car images,
testini göstereceğim.
we can all understand.
resmiyle başlıyoruz ve
about 1.5 million car images,
that can split them into the angle
so I have to start from scratch.
areas of structure in these images.
olarak algılayabilir.
and the computer can now work together.
artık ortak çalışabilir.
about areas of interest
to try and use to improve its algorithm.
are in 16,000-dimensional space,
bilgisayar burada
rotating this through that space,
point out the areas that are interesting.
successfully found areas,
bir şekilde alanları buldu.
the computer more and more
we're looking for.
areas of pathosis, for example,
gibi düşünebilirsin
potentially troublesome nodules.
difficult for the algorithm.
of the cars are all mixed up.
arkaları karmakarışık.
as opposed to the backs,
that this is a type of group
we skip over a little bit,
biraz atlıyoruz bunu,
machine learning algorithm
some of these pictures out,
başladığını görebilirsiniz
how to understand some of these itself.
farkettiğini gösteriyor.
of similar images,
konseptini kullanabiliriz
kullanarak bilgisayarın
entirely find just the fronts of cars.
can tell the computer,
a good job of that.
to separate out groups.
biraz çevirmesine izin
computer try to rotate this for a while,
and the right sides pictures
olarak bulabiliriz.
ipuçları verebiliriz,
the computer some hints,
a projection that separates out
öğrenme algoritmasını
as much as possible
bir projeksiyon bulmaya çalış. "
ah, okay, it's been successful.
ayırabilmek için gereken
of thinking about these objects
is being replaced by a computer,
yerine bilgisayarın geçmesi
something that used to take a team
senede yapacağı bir şeyi,
that takes 15 minutes
5 yineleme içeriyor.
four or five iterations.
classified correctly.
can start to quite quickly
that there's no mistakes.
bilgisayara bildirebiliriz.
let the computer know about them.
bir süreç kullanarak
for each of the different groups,
an 80 percent success rate
that aren't classified correctly,
to 97 percent classification rates.
could allow us to fix a major problem,
olmadığı yerlerde.
of medical expertise in the world.
that there's between a 10x and a 20x
kıtlığı 10 ile 20 katı
in the developing world,
to fix that problem.
yaklaşımını kullanarak
enhance their efficiency
yardım edebilsek?
about the opportunities.
problemlerle de ilgileniyorum.
every area in blue on this map
are over 80 percent of employment.
computers have just learned how to do.
istihdamın %80 'inin
in the developed world
have just learned how to do.
Tamam, bir şey olmayacak.
They'll be replaced by other jobs.
more jobs for data scientists.
için daha fazla iş olacak.
very long to build these things.
were all built by the same guy.
it's all happened before,
of when new things come along
bu şekilde kademeli olarak
grows at this gradual rate,
derin öğrenmeye
büyüme kapasitesine
in capability exponentially.
şeyleri görüyoruz ve:
are still pretty dumb." Right?
computers will be off this chart.
about this capability right now.
düşünmeye başlamalıyız.
in capability thanks to engines.
that after a while, things flattened out.
to generate power in all the situations,
Sanayi devriminden
from the Industrial Revolution,
it never settles down.
asla durulmayacak.
aktiviteler de iyiye gittikçe,
at intellectual activities,
to be better at intellectual capabilities,
never experienced before,
of what's possible is different.
as capital productivity has increased,
hatta biraz inişteydi.
in fact even a little bit down.
having this discussion now.
about this situation,
oldukça ilgisiz olabiliyor.
they don't understand poetry,
olarak anlamıyoruz.
insanların ücret alarak
of their time being paid to do,
harcadıkları şeyleri yapabiliyor.
yapımızı nasıl uyduracağımızı
social structures and economic structures
ve bu yeni gerçekliğin
ABOUT THE SPEAKER
Jeremy Howard - Data scientistJeremy Howard imagines how advanced machine learning can improve our lives.
Why you should listen
Jeremy Howard is the CEO of Enlitic, an advanced machine learning company in San Francisco. Previously, he was the president and chief scientist at Kaggle, a community and competition platform of over 200,000 data scientists. Howard is a faculty member at Singularity University, where he teaches data science. He is also a Young Global Leader with the World Economic Forum, and spoke at the World Economic Forum Annual Meeting 2014 on "Jobs for the Machines."
Howard advised Khosla Ventures as their Data Strategist, identifying opportunities for investing in data-driven startups and mentoring portfolio companies to build data-driven businesses. He was the founding CEO of two successful Australian startups, FastMail and Optimal Decisions Group.
Jeremy Howard | Speaker | TED.com