The Brains Behind AI is a 10-part video series that provides a behind-the-scenes look at some of Canada’s most talented researchers in artificial intelligence, what areas of research they are focusing on over the next five years as Canada CIFAR AI Chairs and CIFAR fellows, and what motivates them to pursue research in a variety of areas that will transform the way we live and work.
CIFAR is leading the Government of Canada’s $125 million Pan-Canadian Artificial Intelligence Strategy, working in partnership with three newly established AI Institutes – Amii in Edmonton, Mila in Montreal and the Vector Institute in Toronto.
Check back weekly for the latest Brains Behind AI video and join the conversation on Twitter using #RealBrains with @CIFAR_News.
I’m Alisa Strome I’m the associate vice
president research and the executive
director of the pan-canadian
AI strategy at Seaford so one of the key
elements of the pan-canadian AI strategy
is the Canada CFRA I chairs program
which is designed to both recruit and
retain some of the world’s great
researchers in AI in Canada as part of
our national ai’s strategy we’re
investing very deeply in fundamental AI
research and the goal is to really
support the scientists who are
developing and advancing the science of
AI and thinking about what are the
opportunities for its applications and
how do we make those better more
efficient more effective one of the
biggest challenges in AI in Canada and
around the world is around equity
diversity and inclusion right now
internationally around the world only
18% of researchers in AI are women and
clearly that’s a problem and I think
it’s a problem that we need to solve
it’s a problem in Canada just like it is
everywhere else in the world and so it’s
see far we’re actually really focusing
very distinctly on that issue many of
the training programs that we offer for
graduate students and even high school
students and Beyond across the country
every summer are really focused on
engaging and working with
underrepresented groups in AI a
particular focus on engaging girls and
women and giving them skills and
expertise so that they can
to pursue careers in AI and help us to
drive this technology forward as a
society if we are going to be able to
really truly benefit on what AI can
offer to humanity we need to make sure
that many many different perspectives
are included
What is Canada doing to support equity, diversity and inclusion in AI research and development? Dr. Elissa Strome is AVP Research and the Executive Director for the Pan-Canadian AI Strategy at CIFAR. She’s one of the brains behind the CIFAR Pan-Canadian AI Strategy. It is the world’s first national AI Strategy, supporting better equity, diversity and inclusion in AI research, training for the next generation, and in the development of future products and services.
I’m Jimmy Paul I’m a assistant professor
at the University of Toronto and also a
faculty affiliate at vector Institute
I’m interested in how to make new and I
was to learn faster because right now
audioletter kind of silly compared to
humans
then with quite a lot of data and a lot
of compute so my research focus or do I
make them faster and how can I
understand the current algorithms that
we apply in deep learning that makes
them work so to me my personal interest
is by building faster learning machines
we can probably tap in some new insight
about how human intelligence works
behind the same right the kind of take a
little peek behind the curtains to
understand what makes us so special to
understand the world around us what gave
us ability to learn the included physics
models and allows us to do like
high-level reasonings about our
long-term plans and short-term plans my
personal belief is that the more
dialogues we have across the disciplines
across the field the better we’ll
understand the problem as a whole of how
should we decide as a society to unable
the technology for everyone for me it’s
a great opportunities to work on very
foundational research problems where
typically the industry search are very
short term and you’re cited in a sense
that we think about two or three years
research program that potentially has a
product on the line where see far to
give me the freedom and able me to focus
on more much more long term research
problems for example AI safety program
and building model-based reinforcement
learning or trying to understand the
theoretical properties of the deep
learning when we use today
you
How do we make neural networks faster? Canada CIFAR AI Chair Jimmy Ba is the brains behind the Adam Optimizer, one of the go-to algorithms to train deep learning models. A faculty member of the Vector Institute in Toronto and Assistant Professor in the Department of Computer Science at the University of Toronto, he’s leading research in AI to develop better technologies for all.
Can we train computers to understand language the way humans do? And can we use this language to build a virtual 3-D world? Canada CIFAR AI Chair Angel Chang is using AI to convert natural language into 3-D representations of the world. This will better prepare the systems of the future to interpret and act in the real world. Angel is a member of the Alberta Machine Intelligence Institute (Amii) and an assistant professor at Simon Fraser University.
my name is Jakob NGO and I’m a professor
at University Maria I am the scientific
director of miele my collaborators and I
are trying to understand the limitations
of guarantee.i which is using machine
learning a lot of deep learning we are
very interested in the perspective of
how agents like animals and humans who
act in their environment actually have
some understanding of their environment
understand that the causal relationships
in it and how they can reason in it by
doing things that current learning based
systems are not really yet good at but
that humans are very good at the reason
I got into AI when I was in my early 20s
is because I thought that intelligence
was one of the biggest mystery that
science needed to understand and in
addition the reason I got into this
particular approach to AI base on neural
networks is because it sits at the
intersection of the question of
understanding human and animal
intelligence and understanding how to
build intelligent machines and it’s it
rests on the idea that maybe the same
principles would apply to both so the
sea far AI chair program is very
important for Canada’s strategic
development both economically and
socially the chairs are bringing
researchers to Canada are keeping good
researchers in Canada both professors
but also the students and postdocs that
they hire and they’re also showing to
the rest of the world that we are
serious about AI that we are investing
in AI and that it’s a long-term project
for the country
Could AI solve the mystery of human and animal intelligence? Canada CIFAR AI Chair Yoshua Bengio is pioneering research in deep learning and is one of the brains behind artificial neural networks, an approach that teaches computers to mimic human intelligence. Yoshua is co-director and fellow of CIFAR’s Learning in Machines & Brains program a professor at Université de Montréal, founder and scientific director of Mila, scientific director of IVADO. He is a Canada Research Chair in Statistical Learning Algorithms and a recipient of the 2018 A.M Turing Award, often referred to as the “Nobel Prize” in computing for his contributions to deep learning.
my name is Martha white I’m an assistant
professor at the University of Alberta
and computing science department and
also the director of RLE I I work in an
area called reinforcement learning and
it’s about having an agent interact with
the world the real world around it in a
trial-and-error way and the higher level
goal is about having an agent learning
in a continual way you want you don’t
just want to have an ego that you
programmed offline that you’re deploying
to the real world we kind of think of
them like agents like we are they are
constantly learning as they’re
interacting with the world so the
potential of applications is really big
I would argue that anywhere where humans
are currently doing intelligent things
or that we have intelligent systems is a
great place for reinforcement learning
when you have this agent and it’s
continually interacting in the world it
needs to actually take in that flow of
data and summarize it into some compact
form that lets it reason about the world
so you can’t just remember everything
it’s seen you an imagined it has to pick
out the most important things and store
that information and unconstant Lee
update what it currently believes about
the world so research in general is it’s
always been important but I would argue
it’s becoming even more and more
important as you’re entering this age
where information is just a huge economy
you know expertise is extremely
important if from an industry
perspective I think investing in
research is great because we help
develop that industry but we also
attract really good people to Canada who
will then produce good students that
will start those companies or they
themselves will decide to help start
those companies and that’s in fact
exactly what’s happened in Edmonton
How do we teach machines to make intelligent decisions in the real world? Canada CIFAR AI Chair Martha White is using reinforcement learning techniques to train autonomous agents to make smarter and more efficient decisions through trial and error. The applications are wide-ranging and could be used in autonomous vehicles and robotics, as well as industries such as finance, manufacturing and more. .Martha White is a member of Amii, an assistant professor at the University of Alberta’s Department of Computing Science, and the director of RLAI.
my name is Blake Richards I’m a
neuroscientist and artificial
intelligence researcher I’m an assistant
professor at McGill University my
research is about trying to uncover
general principles of intelligence that
apply equally to both machines and
people the kinds of things that we study
are for example you know what do you
actually use your memory for and how
does your brain decide what to remember
and what to forget that’s actually a
really fundamental principle to your own
intelligence and your own understanding
of the world you can imagine that if you
were able to build a computer system
that remembered in the way that humans
do it would be just that much more
intuitive and potentially able to take
on human roles in society as opposed to
just being some kind of alien black box
that we stick stuff on see far has had a
big impact on my direction as a
researcher because what it really allows
us to do as researchers is to explore
the edges of our knowledge but also
because a big part of their mission is
that you have to come and talk to other
researchers from other disciplines you
have to get other perspectives come to
lots of meetings think about things in
ways that you hadn’t thought about
before and interact with people you
wouldn’t normally interact with
How does your brain determine what’s important enough to remember? Canada CIFAR AI Chair and CIFAR Fellow Blake Richards is using artificial intelligence to better understand the fundamental principles of intelligence for both machines and brains. Blake is a fellow in the CIFAR Learning in Machines & Brains program, a member of Mila and professor at the Montréal Neurological Institute and the School of Computer Science at McGill University.
I’m Alanna fish I’m an assistant
professor at the University of Alberta
and I’m also a Canada Safari i J
my research is trying to find
connections between the way computers
understand the world and the way we
understand the world so when we build
computer models of for example computer
vision we ask computers to tell us
what’s in a picture and it turns out the
way they learn to do that actually looks
a lot like the way the human brain does
it so I’m my research is trying to draw
those connections between the way the
brain works and the way that machine
learning models work we understood how
people perceived images we could build
computer models that perceive images in
the same way and that might mean they’re
better at that perception it could also
mean things like when they make a
mistake it’s a more human mistake and
that sort of thing could be really
important for applications where
computers are interacting with the real
world I think Canada has been so well
positioned for the AI revolution I think
we happen to have a really strong group
of people already here and then with the
funding from Seifer on the Canadian
government were able to grow that team
in a way that has really put us ahead of
other countries I think it’s a great
time to be in Canada for AI but the
thing that I think has been most
important to me with my involvement with
see far has been the networking and the
getting to know other people who are
working on AI so that has been has had a
huge impact on my career a measurable
impact on my career already and I’m
quite a new researcher so having those
sorts of opportunities early on is
really important you can never really
tell what what new science is going to
change the world and so finding the
right people smart people and giving
them the money they need to make
progress is really important and it’s
what will make science into reality
When computers see an image, how do they perceive them? Canada CIFAR AI Chair Alona Fyshe is pioneering research in computer vision and helping computers to see the world the way humans do. Alona is a fellow in the CIFAR Child & Brain Development program, an assistant professor at the University of Alberta and a member of Amii.
Neural networks have revolutionized areas such as computer vision and natural language processing but how do we train them to be faster and more efficient? Canada CIFAR AI Chair Roger Grosse is helping machines to better predict and adapt to different situations. Roger is an assistant professor of computer science at the University of Toronto, a founding member of the Vector Institute, a Canada Research Chair in Probabilistic Inference and Deep Learning and co-creator of Metacademy.
The AI systems of the future, such as robots and autonomous vehicles, will play a more integrated role in society. Canada CIFAR AI Chair Graham Taylor is exploring how to make machines more accurate and robust so that they can better interact with humans in the real world. Graham is a fellow in the CIFAR Learning in Machines & Brains program, a CIFAR Azrieli Global Scholar, a faculty member at the Vector Institute, an associate professor of engineering at the University of Guelph and an academic director at Next AI.
Solving the mysteries of machine intelligence could lead to more profound answers about our own human intelligence. Rich Sutton is pioneering the field of reinforcement learning, a type of machine learning that allows machines to learn from interactions with its environment. RIch is a senior fellow at CIFAR, a fellow at Amii, a professor at the University of Alberta and founder of DeepMind.
Machines learn best when they have lots of data, but large datasets are not always easy to come by. Canada CIFAR AI Chair Hugo Larochelle is advancing research in few shot learning, a technique commonly employed in computer vision. Hugo is looking at ways to make systems faster and more accurate in the absence of lots of training data, a feat that could make machine learning more accessible for broader audiences. Hugo is an associate fellow in the CIFAR Learning in Machines & Brains program, a faculty member of Mila, a professor at the Université de Montréal. He leads the Google Brain group in Montréal.
Canada CIFAR AI Chair James Wright is using machine learning algorithms to predict human strategic behaviour, an approach that seeks to find ways to align interests and goals between many people, things and situations. His research will have considerable impact to the areas of economics, commerce and market. It could revolutionize how we organize online platforms to benefit all. James is a fellow at Amii, an assistant professor at the University of Alberta.
This website stores cookies on your computer. These cookies are used to collect information about how you interact with our website and allow us to remember you. We use this information in order to improve and customize your browsing experience and for analytics and metrics about our visitors both on this website and other media. To find out more about the cookies we use, see our Privacy Policy.
