Blaise Agüera y Arcas: How computers are learning to be creative
بليز أوغيرا إي أركاس: كيف تتعلم الحواسيب لتصبح مبدعة؟
Blaise Agüera y Arcas works on machine learning at Google. Previously a Distinguished Engineer at Microsoft, he has worked on augmented reality, mapping, wearable computing and natural user interfaces. Full bio
Double-click the English transcript below to play the video.
that works on machine intelligence;
يعمل في مجال الذكاء الاصطناعي؛
of making computers and devices
النظام الهندسي لصنع الحواسيب والأجهزة
that brains do.
التي يفعلها الدماغ.
interested in real brains
in the things that our brains do
بالأمور التي تقوم بها أدمغتنا
to the performance of computers.
علي أداء الحواسيب.
has been perception,
out there in the world --
يمكن للأشياء المحيطة --
for example, that our team makes,
على سبيل المثال، التي يصنعها فريقنا،
on Google Photos to become searchable,
قابلة للبحث،
out there into the world.
ملموس يهم العالم.
our work on machine perception
فإن عملنا في الإدراك الإصطناعي
with the world of machine creativity
بالإبداع الآلي
had a penetrating insight
بصيرةً نافذة
between perception and creativity.
بين الإدراك والإبداع.
has a statue inside of it,
is to discover it."
Michelangelo was getting at
is an act of imagination
and perceiving and imagining,
بكل التفكير والإستيعاب والتخيّل،
with a brief bit of history
the heart or the intestines,
about a brain by just looking at it,
بمجرد النظر اليه،
الذين نظروا في الدماغ
of this thing all kinds of fanciful names,
المصطلحات المبهرجة،
doesn't tell us very much
لا يخبرنا بالكثير
developed some kind of insight
نوع من البصيرة
Santiago Ramón y Cajal,
(سانتياغو رامون كاخال)،
or render in very high contrast
المفردة في الدماغ
their morphologies.
that he made of neurons
الخلايا العصبية
of different sorts of cells,
أنواع الخلايا،
was quite new at this point.
في تلك المرحلة.
النهائية،
طويلة جداً جداً
very, very long distances --
to some people in the 19th century;
were just getting underway.
لا تزال قيد البناء.
of Ramón y Cajal's, like this one,
كهذه الرسمة،
that Ramón y Cajal started.
(رامون كاخال).
of Neuroscience.
is about one cubic millimeter in size,
مكعب واحد،
very small piece of it here.
حوالي مايكرون واحد.
(ميتوكوندريا)
tiny block of tissue.
of hair is about 100 microns.
much, much smaller
electron microscopy slices,
المأخوذة بمجهر إلكتروني تسلسلي،
in 3D of neurons that look like these.
نموذج ثلاثي الأبعاد لخلية عصبية تبدو كهذه.
style as Ramón y Cajal.
لدى (رامون كاخال).
be able to see anything here.
one neuron to another.
ببعضها البعض.
ahead of his time,
لعصره بعض الشيء،
over the next few decades.
استخدمت الكهرباء،
تقنياتنا على نحو كافٍ
was advanced enough
experiments on live neurons
خلايا عصبية حية
when computers were being invented,
اختراع الحواسيب.
of modeling the brain --
الدماغ --
as Alan Turing called it,
عليها (ألان تورينغ)،
looked at Ramón y Cajal's drawing
رسومات (رامون كاخال)
imagery that comes from the eye.
الصور القادمة من العين.
like a circuit diagram.
دارة كهربائية.
in McCulloch and Pitts's circuit diagram
الدارة لكل من (ماكولوتش) و(بيتس)
of computational elements
العناصر الحاسوبية
one to the next in a cascade,
بتسلسل،
visual information would need to do.
المعلومات البصرية.
for us to do with our brains.
باستخدام أدمغتنا.
that for a computer,
أنه بالنسبة لحاسوب،
just a few years ago.
بضعة سنوات قليلة.
this task is easy to do.
بسهولة.
and the word "bird,"
connected to each other
المرتبطة ببعضها
ضمن القشرة الدماغية البصرية،
inside our visual cortices,
to have the capability
on the computer.
that actually looks like.
about as a first layer of neurons,
العصبونات،
how it works in the eye --
after another layer of neurons,
من العصبونات،
of different weights.
مختلفة.
of all of those synapses.
properties of this network.
or a small group of neurons
العصبونات
those three things --
in the neural network,
العصبية،
these synapses in the neural network.
العصبية.
is just a simple formula,
بسيطة،
going on there, of course,
of mathematical operations.
الرياضية.
that if you have one equation,
لدينا معادلة واحدة،
by knowing the other two things.
المتغيرين الآخرين.
that the picture of a bird is a bird,
and w and x are known.
معلومان.
you know the pixels.
البيكسلات.
a relatively straightforward problem.
and you're done.
doing exactly that.
on a mobile phone,
amazing in its own right,
billions and trillions of operations
بل ترليونات العمليات
picture of a bird,
"Yes, it's a bird,"
with a network of this sort.
and the y is the unknown.
difficult part, of course,
do we figure out the w,
of solving for w,
with the simple equation
المعادلة البسيطة
it's the inverse to multiplication,
very non-linear operation;
عملية غير خطية
to solve the equation
is fairly straightforward.
نوعاً ما.
a little algebra trick,
to the right-hand side of the equation.
من المعادلة.
about it as an error.
بالطريقة الصحيحة،
for w the right way,
to minimize the error,
computers are very good at.
أجهزة الحاسوب جداً.
sort of play Marco Polo,
يلعب (ماركو بولو)،
successive approximations to w.
تقريبة متعاقبة ل w.
but after about a dozen steps,
لكن بعد الكثير من الخطوات،
which is close enough.
وهي قيمة تقريبية كافية.
a lot of known x's and known y's
المعلومة
through an iterative process.
that we do our own learning.
this is not a bird."
ليس هذا طائراً."
for those neural connections.
تلك الوصلات العصبية.
x and w fixed to solve for y;
لكي نحل y ،
Alex Mordvintsev, on our team,
فريقنا (أليكس ماريفينسف)،
with what happens if we try solving for x,
حل المعادلة لأجل x،
that you've trained on birds,
على أن الذي أمامك طائر،
the same error-minimization procedure,
تقليل الخطأ،
trained to recognize birds,
على التعرف على الطيور،
generated entirely by a neural network
بواسطة شبكة عصبية
rather than solving for y,
by Mike Tyka in our group,
of William Kentridge's artworks,
الفنية،
over the space of different animals,
متباينة من الحيوانات،
to recognize and distinguish
morph from one animal to another.
من حيوان لآخر،
have tried reducing
out of the space of all things
من مساحة كل الأشياء
over that entire surface,
you make a kind of map --
خريطة نوعاً ما --
the network knows how to recognize.
التي تستطيع الشبكة تمييزها.
"armadillo" is right in that spot.
في البقعة المناسبة.
of networks as well.
من الشبكات أيضاً.
to recognize faces,
in a y that says, "me,"
psychedelic picture of me
multiple points of view at once
نظر في نفس الوقت
to get rid of the ambiguity
لكي تتخلص من الغموض
or another pose,
another kind of lighting.
this sort of reconstruction,
of different points of view,
من نقاط مختلفة،
his own face as a guide image
وجهه كصورة دليلية
to reconstruct my own face.
that optimization process.
more like a coherent face,
يشبه الوجه المتماسك،
with a blank canvas
that is itself already some other image.
والتي هي نفسها صورة ما مسبقاً.
that is designed to categorize
man-made structures, animals ...
أشخاص، أشكال، حيوانات ...
with just a picture of clouds,
what it sees in the clouds.
بين الغيوم.
you spend looking at this,
will see in the clouds.
to hallucinate into this,
لكي تهلوس الى هذا،
zooms hallucinates, zooms.
يهلوسها ويكبرها.
of the network, I suppose,
لهذه الشبكة كما أعتقد،
is eating its own tail.
What do I think I see next?"
مالذي أعتقد أني سأراه بعدها؟
called "Higher Education" --
تحت عوان (التعليم العالي) --
marijuana was legalized.
is not constrained.
because they're really fun to look at.
لأنه من الممتع النظر اليها
a camera that takes a picture,
writes a poem using neural networks,
قصيدة باستخدام الشبكات العصبية،
has been trained
are very intimately connected.
things in the world,
في هذا العالم،
المعاكس، لتولد.
Michelangelo really did see
any being, any alien
perceptual acts of that sort
machinery that's used in both cases.
and creativity are by no means
لا يعنيان بالضرورة
that can do exactly these sorts of things.
بنفس هذه الأشياء.
the brain is computational.
فالدماغ في الأساس حسابي.
in designing intelligent machinery.
of those early pioneers,
is not just about accounting
we modeled them after our minds.
to understand our own minds better
فهماً أفضل
ABOUT THE SPEAKER
Blaise Agüera y Arcas - Software architectBlaise Agüera y Arcas works on machine learning at Google. Previously a Distinguished Engineer at Microsoft, he has worked on augmented reality, mapping, wearable computing and natural user interfaces.
Why you should listen
Blaise Agüera y Arcas is principal scientist at Google, where he leads a team working on machine intelligence for mobile devices. His group works extensively with deep neural nets for machine perception and distributed learning, and it also investigates so-called "connectomics" research, assessing maps of connections within the brain.
Agüera y Arcas' background is as multidimensional as the visions he helps create. In the 1990s, he authored patents on both video compression and 3D visualization techniques, and in 2001, he made an influential computational discovery that cast doubt on Gutenberg's role as the father of movable type.
He also created Seadragon (acquired by Microsoft in 2006), the visualization technology that gives Photosynth its amazingly smooth digital rendering and zoom capabilities. Photosynth itself is a vastly powerful piece of software capable of taking a wide variety of images, analyzing them for similarities, and grafting them together into an interactive three-dimensional space. This seamless patchwork of images can be viewed via multiple angles and magnifications, allowing us to look around corners or “fly” in for a (much) closer look. Simply put, it could utterly transform the way we experience digital images.
He joined Microsoft when Seadragon was acquired by Live Labs in 2006. Shortly after the acquisition of Seadragon, Agüera y Arcas directed his team in a collaboration with Microsoft Research and the University of Washington, leading to the first public previews of Photosynth several months later. His TED Talk on Seadragon and Photosynth in 2007 is rated one of TED's "most jaw-dropping." He returned to TED in 2010 to demo Bing’s augmented reality maps.
Fun fact: According to the author, Agüera y Arcas is the inspiration for the character Elgin in the 2012 best-selling novel Where'd You Go, Bernadette?
Blaise Agüera y Arcas | Speaker | TED.com