Jeremy Howard: The wonderful and terrifying implications of computers that can learn
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,
الحاسب القيام بمهمة ما
that haven't done it yourself,
the computer to do
that you don't know how to do yourself,
لاتعرف كيفية القيام به بنفسك،
to be a great challenge.
by this man, Arthur Samuel.
ارثر صامويل
ان يشتري هذا الحاسوب
لعبة الشطرنج
how to be better than you at checkers?
يمكن للحاسوب ان يكون أفضل منك في الشطرنج؟
against itself thousands of times
and in fact, by 1962,
the Connecticut state champion.
the father of machine learning,
learning practitioner.
machine learning practictioners.
previously unsolved problems,
مشاكل لم تحل من قبل
hundreds of times.
I was able to find out
can do in the past, can do today,
فعله في الماضي وما يستطيعه اليوم
machine learning commercially was Google.
للتعلم الآلي هو جوجل
possible to find information
on machine learning.
commercial successes of machine learning.
تجارية كبيرة في التعلم الآلي
products that you might like to buy,
التي قد تفضل شرائها
who your friends might be
the power of machine learning.
learned how to do this from data
كيف تفعل هذا من البيانات
the two world champions at "Jeopardy,"
يهزم بطلي عالم في مسابقة "جابردي"
and complex questions like this one.
ومعقدة مثل
from this city's national museum in 2003
(along with a lot of other stuff)"]
to see the first self-driving cars.
the difference between, say,
well, that's pretty important.
those programs by hand,
this is now possible.
over a million miles
يتعلم كيفية فعل أشياء
don't know how to do ourselves,
I've seen of machine learning
called Geoffrey Hinton
جوفري هينتون
automatic drug discovery.
is not just that they beat
or the international academic community,
ميرك أو المجتمع الأكاديمي الدولي
in chemistry or biology or life sciences,
علوم الحياة أو الكيمياء أو الأحياء
called deep learning.
فائقة تدعى التعلم العميق
the success was covered
article a few weeks later.
لنيويورك تايمز منذ عدة أسابيع
here on the left-hand side.
inspired by how the human brain works,
تم إستلهامها من كيفية عمل العقل البشري
on what it can do.
computation time you give it,
showed in this article
result of deep learning
can listen and understand.
أخر خطوة
to take in this process
of information from many Chinese speakers
المعلومات من العديد من متحدثي الصينية
and converts it into Chinese language,
an hour or so of my own voice
so that it would sound like me.
المكتوب لحديث حتى يبدو كصوتي
أنجازه في هذا المجال
a machine learning conference in China.
في مؤتمر التعلم الآلي بالصين
at academic conferences
بالمؤتمرات الأكاديمية
at TEDx conferences, feel free.
بمؤتمرات TEDx ، تصرفوا بحريتكم.
was happening with deep learning.
شكراً لكم
was deep learning.
كان التعلم العميق
in the top right, deep learning,
اليمين كان التعلم العميق
was deep learning as well.
التعلم العميق أيضاً
this extraordinary thing.
can seem to do almost anything,
تقريباً أن تفعل أي شئ
it had also learned to see.
أيضاً قد تعلمت أت ترى.
Recognition Benchmark,
to recognize traffic signs like this one.
على إشارات المرور مثل هذه
recognize the traffic signs
it was better than people,
أنها أفضل من البشر
better than people.
they had a deep learning algorithm
أن لديهم خوارزمة تعلم عميق
on 16,000 computers for a month,
about concepts such as people and cats
عن مفاهيم مثل التاس والقطط
that humans learn.
by being told what they see,
إخبارهم عن ما يروه
what these things are.
who we saw earlier,
الذي رأيناه منذ قليل
from one and a half million images
لإكتشاف من بين مليون ونصف صورة
to a six percent error rate
قلصنا نسبة الخطأ لـ 6%
an extraordinarily good job of this,
تستطيع القيام بأعمال فائقة
location in France in two hours,
that they fed street view images
to recognize and read street numbers.
على وقراءة أرقام الشوارع
it would have taken before:
the Chinese Google, I guess,
to Baidu's deep learning system,
نظام بيدو للتعلم العميق
has understood what that picture is
فهم ماذا تكون هذه الصورة
have similar backgrounds,
at the text of a web page.
عن نص مكتوب على صفحة ويب
really understand what they see
كمبيوتر تستطيع بالفعل فهم ما تراه
في قواعد بيانات
of images in real time.
now that computers can see?
أجهزة الكمبيوتر أن ترى؟
that computers can see.
تستطيع الرؤية
has done more than that.
with deep learning algorithms.
showing the red dot at the top
is expressing negative sentiment.
is near human performance
أصبح قريب من الأداء البشري
and what it is saying about those things.
وما تقوله عن هذه الأشياء.
been used to read Chinese,
الصينية
Chinese speaker level.
out of Switzerland
or understand any Chinese.
in the world for this,
human understanding.
put together at my company
all this stuff together.
have no text attached,
these pictures
to the text that I'm writing.
understanding my sentences
something like this on Google,
and it will show you pictures,
searching the webpage for the text.
عن نصوص بصفحة الموقع
understanding the images.
have only been able to do
can not only see but they can also read,
ليس فقط الرؤية بل القراءة أيضاً
can understand what they hear.
I'm going to tell you they can write.
أخبرك أنها تستطيع الكتابة.
using a deep learning algorithm yesterday.
out of Stanford generated.
خوارزمة بستانفورد
to describe each of those pictures.
لوصف كلاً من هذه الصور
a man in a black shirt playing a guitar.
يرتدي قميص أسود ويعزف على الجيتار
it's seen black before,
this novel description of this picture.
الجديد عن هذه الصورة
performance here, but we're close.
البشري ولكننا قريبين
the computer-generated caption
الشرح الذي أنتجه الكمبيوتر
well past human performance
الكمبيوتر الأداء البشري
to very exciting opportunities.
that they had discovered
make a prognosis of a cancer.
علي التنبؤ بالسرطان
looking at tissues under magnification,
a machine learning-based system
than human pathologists
البشريين في علوم الأمراض
for cancer sufferers.
were the predictions more accurate,
that humans can understand.
جديدة يستطيع البشر إدراكها
that the cells around the cancer
بالفعل أن الخلايا حول السرطان
the cancer cells themselves
had been taught for decades.
الأمراض لعشرات السنين
they were systems developed
and machine learning experts,
كلاً من الطب والتعلم الألي
we're now beyond that too.
فقد تخطينا هذا الأن أيضاً
identifying cancerous areas
can identify those areas more accurately,
يستطيع تحديد هذه المناطق بدقة أكبر
as human pathologists,
using no medical expertise
بدون إستخدام أي خبرة طبية
no background in the field.
عن هذا المجال
about as accurately as humans can,
العصبية بنفس دقة البشر
with deep learning
background in medicine.
no previous background in medicine,
لديه أي خلفية طبية سابقة
to start a new medical company,
that it ought to be possible
using just these data analytic techniques.
فقط أساليب تحليل البيانات
has been fantastic,
but from the medical community,
ولكن من المجتمع الطبي
the middle part of the medical process
نأخذ الجزء الأوسط من العملية الطبية
as much as possible,
what they're best at.
to generate a new medical diagnostic test
لإستخراج إختبار تشخيص طبي جديد
three minutes by cutting some pieces out.
بإقتطاع بعض الأجزاء منها
creating a medical diagnostic test,
عمل إختبار تشخيص طبي
a diagnostic test of car images,
we can all understand.
about 1.5 million car images,
that can split them into the angle
تقسيم الصور للزاوية
so I have to start from scratch.
تماماً لذا كا ن علي البدء من نقطة الصفر
areas of structure in these images.
على مناطق بنية هذه الصور
and the computer can now work together.
يستطيعان الأن العمل سوياً
about areas of interest
to try and use to improve its algorithm.
ليطور بعد ذلك من خوارزمته
are in 16,000-dimensional space,
بالفعل بفراغ ذو 16000 بعد
rotating this through that space,
الصور في هذا الفراغ
point out the areas that are interesting.
يشير إلى المناطق التي تهمه
successfully found areas,
هذه المناطق بنجاح
the computer more and more
we're looking for.
التشخيص الطبي
areas of pathosis, for example,
potentially troublesome nodules.
difficult for the algorithm.
of the cars are all mixed up.
as opposed to the backs,
عكس الخلفيات
that this is a type of group
المجموعه
we skip over a little bit,
لأننا نتخطاه قليلاً
machine learning algorithm
خوارزمة التعلم الآلي
some of these pictures out,
تغذية بعض هذه الصور
how to understand some of these itself.
على كيفية فهم البعض بنفسها
of similar images,
entirely find just the fronts of cars.
يجد فقط مقدمات السيارات
can tell the computer,
a good job of that.
to separate out groups.
computer try to rotate this for a while,
يحاول إدارتها لوقت
and the right sides pictures
the computer some hints,
a projection that separates out
as much as possible
ah, okay, it's been successful.
of thinking about these objects
للتفكير في هذه الأشياء
is being replaced by a computer,
الكمبيوتر مكان البشر
something that used to take a team
إعتدنا أن ينفذه فريق
that takes 15 minutes
four or five iterations.
classified correctly.
can start to quite quickly
that there's no mistakes.
let the computer know about them.
for each of the different groups,
an 80 percent success rate
that aren't classified correctly,
لم يتم تصنيفه صحيحاً
to 97 percent classification rates.
could allow us to fix a major problem,
of medical expertise in the world.
that there's between a 10x and a 20x
in the developing world,
to fix that problem.
enhance their efficiency
about the opportunities.
every area in blue on this map
are over 80 percent of employment.
أكثر من 80% من التشغيل
computers have just learned how to do.
تعلم الكمبيوتر مؤخرا كيف يفعلها
in the developed world
have just learned how to do.
الكمبيوتر كيفية عملها
They'll be replaced by other jobs.
more jobs for data scientists.
very long to build these things.
وقت طويل لبناء هذه الأشياء
were all built by the same guy.
it's all happened before,
of when new things come along
عندما حدثت أشياء جديدة
grows at this gradual rate,
in capability exponentially.
are still pretty dumb." Right?
computers will be off this chart.
الكمبيوتر خارج هذه الخريطة
about this capability right now.
in capability thanks to engines.
that after a while, things flattened out.
to generate power in all the situations,
لتوليد الكهرباء بجميع المواقف
from the Industrial Revolution,
it never settles down.
at intellectual activities,
to be better at intellectual capabilities,
أفضل لتطوير مستويات ذكائهم
never experienced before,
of what's possible is different.
as capital productivity has increased,
زادت إنتاجية رأس المال
in fact even a little bit down.
having this discussion now.
about this situation,
they don't understand poetry,
of their time being paid to do,
social structures and economic structures
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