hi everyone my name is deshita and I’m a
researcher at the human-computer
interaction lab at MIT
my background is an architecture and
computer science and my passion is to
build machines that can help designers
with creative problem-solving the
motivation for my research comes from my
personal experience as an architect I
was never the lone creative genius
being a designer for the last 10 years
has made me realize that creative
thinking is a social exercise every time
I struggled with a problem that required
a creative solution that out-of-the-box
thinking I seeked perspectives of those
around me and somehow they our way of
looking at the problem differently than
me opened up so many avenues of
solutions for me the environments we
thrive in the context and cultures we
engage with but most importantly the
people we interact with directly
influence our creative abilities so when
I came to MIT I started researching on
building machines that could be creative
partners I wanted to understand why and
how did my interactions with others
enabled me to develop creative solutions
as I dug deeper I realized that creative
thinking was not as mysterious as I
believed it to be in fact each one of us
can be creative all the time
let me show an example take a moment and
look at this image what did you just see
more importantly what did you think if
you were a Beatles fan you probably
thought of their Blackbird song if you
were a mother of a toddler you probably
thought of the nursery rhyme for in 20
blackbirds baked in a pie or if you’re a
Game of Thrones fan like me you’re
probably thought of the three-eyed Raven
why we all looked at the same thing we
had our own unique ways of looking at
the blackboard and if you a poet Wallace
Stevens you had not one but 30 in
different ways of looking at the
blackboard
Wallace’s poem highlights the most
extraordinary ability of human minds
creative thinking by merely able to see
different meanings in the same object
and this form of creative thinking is
not just evident in poetry and
storytelling but also in visual art for
example Wallace’s poem inspired artist
Michael Spafford to create his
interpretations of Wallace’s poem as
paintings interestingly ten years later
he revisited his own paintings to form
new expressions our ability to see
different meanings in the objects around
us and then express these visions across
different mediums is what makes us a
creative society creative thinking is
nothing but seeing different meanings
that adds value to our knowledge and
sometimes it’s revisiting and re looking
at our interpretations and forming new
meanings it is as if the world is one
giant painting and we are constantly
playing around with it to discover new
meanings discover hidden connections and
this simple idea is powerful because
what it means is that each one of us can
be creative as long as we can develop
our own unique ways of looking at the
world and this way of looking uniquely
on the world is a term that is
scientifically called
as cognitive diversity as a visual
analogy consider for example you and me
were looking at the same object a prism
based on how we look at it where we are
looking from we would form different
rules to parse this object we would form
different interpretations of this object
and we will embed different meanings in
this object this is what makes this
cognitively diverse from each other
computationally speaking cognitive
diversity is merely being able to use
different perspectives and breaking
objects in different
opponents using different rules and then
using different heuristics to form new
interpretations so if you’re all
cognitively diverse from each other how
does it affect our creative ability this
brings me to the second most
extraordinary ability of human
intelligence social interaction when we
socially interact with each other it’s
as if we exchange our rules of parsing
the world with each other and in the
process we transform each other’s way of
looking at things together we engage not
just in a dialogue but in a creative
thinking process in the end it doesn’t
even matter whether we understood each
other completely or not as long as we
succeeded in transforming our own ways
of looking at the world so the case that
I’m making is that with just these two
abilities one the ability to be
cognitively diverse and second the
ability to socially interact is what
makes us creative Society however we are
spending less time embracing our diverse
thoughts and ideas we are spending less
time problem-solving with humans and
more times with machines and this
machines are programmed to give us the
exact same answer every single time we
query it so this got me thinking can
machines ever learn to be cognitively
diverse can machines ever learn to give
us different perspectives every single
time we ask it the challenge and making
such algorithms is that creative
thinking is neither strictly logical nor
is it entirely random our thinking is so
ephemeral one moment we might look at
something one way another moment
completely differently it is as if we
dynamically decomposed our thoughts fuse
them embed new meanings and recreate new
thoughts it is as if the units of
thoughts are evanescent in nature and
they serve the purpose of explaining the
meaning only momentarily and then
disappear and reappear to make new
meanings this is because we do apply
rules but we are constantly changing
rules to form new interpretations and
then form new rules again
and making machines that can exhibit
such kind of behavior is computationally
challenging given the current frameworks
with machine learning it’s required
using a different paradigm a
computational framework called shape
grammar conceptualized by George Stinney
a professor at MIT provided a way for me
to address this challenge using shape
grammars we can now build machines that
can compute multiple perspectives
consider an example of these three
triangles if I ask people what do they
see most will say three triangles some
however would see two triangles some
would see four triangles somebody like
me would see nine lines and the
possibilities are endless our human mind
is capable of incorporating all these
different answers machines on the other
hand are programmed algorithmic ly to
give the exact same answer three
triangles but what if our frameworks
were not so static and rigid shape
grammars provides us with a framework to
compute all these multiple possibilities
this is because the machine computes
with new set of rules and new components
every single time the asker it can
either have full partial or no memory of
the previous state
so in short using using shape grammars
we can encompass all these indefinite
possibilities and perspectives using a
same simple mechanism but here is the
incredible part this framework can be
applied not just to compute with shapes
and visual forms in my research I use
the principles of shape grammar and
applied it to creative thinking now
imagine that instead of computing which
shapes the machine is now computing with
abstract thoughts and instead of making
rules of placement of shapes the machine
is now making rules of placement of
abstract thoughts this meaningful
arrangement of thoughts or spatial
relations is nothing but ideas so if
thoughts were shapes not
literally and ideas were spatial
relations then creative thinking can be
computed using shape grammar using this
framework we can now build parser
machines that can break complex ideas
into simpler thoughts we can build
interpreter machines that can combine
simpler thoughts and make complex ideas
or we can make inference machines that
can compute with multiple ideas and
start connecting these thoughts and help
us discover meanings which our human
mind was incapable of discovering in
short I’m saying we can build machines
that can think might not think is
logically written still think and if
these machines can learn to share their
rules of parsing with us they can become
true creative partners instead of being
passive interfaces they can now actively
start interacting with us to augment our
creative abilities my vision for the
future of creativity is not where
machines are more creative than us but
one where machines inspire each one of
us to be more diverse in our thinking
more creative in our thinking with
access to wast knowledge an incredible
computing power machines can now provide
us with different diverse perspectives
as we bring in our diverse experiences
to make the sense of the world so let us
build machines that can help us step
outside our own fixed ways of thinking
so that we can understand the world not
in one way or 13 ways but in definite
ways because like Minsky said if we
understood something one way then he
would not have understood it at all
thank you
[Applause]
researcher at the human-computer
interaction lab at MIT
my background is an architecture and
computer science and my passion is to
build machines that can help designers
with creative problem-solving the
motivation for my research comes from my
personal experience as an architect I
was never the lone creative genius
being a designer for the last 10 years
has made me realize that creative
thinking is a social exercise every time
I struggled with a problem that required
a creative solution that out-of-the-box
thinking I seeked perspectives of those
around me and somehow they our way of
looking at the problem differently than
me opened up so many avenues of
solutions for me the environments we
thrive in the context and cultures we
engage with but most importantly the
people we interact with directly
influence our creative abilities so when
I came to MIT I started researching on
building machines that could be creative
partners I wanted to understand why and
how did my interactions with others
enabled me to develop creative solutions
as I dug deeper I realized that creative
thinking was not as mysterious as I
believed it to be in fact each one of us
can be creative all the time
let me show an example take a moment and
look at this image what did you just see
more importantly what did you think if
you were a Beatles fan you probably
thought of their Blackbird song if you
were a mother of a toddler you probably
thought of the nursery rhyme for in 20
blackbirds baked in a pie or if you’re a
Game of Thrones fan like me you’re
probably thought of the three-eyed Raven
why we all looked at the same thing we
had our own unique ways of looking at
the blackboard and if you a poet Wallace
Stevens you had not one but 30 in
different ways of looking at the
blackboard
Wallace’s poem highlights the most
extraordinary ability of human minds
creative thinking by merely able to see
different meanings in the same object
and this form of creative thinking is
not just evident in poetry and
storytelling but also in visual art for
example Wallace’s poem inspired artist
Michael Spafford to create his
interpretations of Wallace’s poem as
paintings interestingly ten years later
he revisited his own paintings to form
new expressions our ability to see
different meanings in the objects around
us and then express these visions across
different mediums is what makes us a
creative society creative thinking is
nothing but seeing different meanings
that adds value to our knowledge and
sometimes it’s revisiting and re looking
at our interpretations and forming new
meanings it is as if the world is one
giant painting and we are constantly
playing around with it to discover new
meanings discover hidden connections and
this simple idea is powerful because
what it means is that each one of us can
be creative as long as we can develop
our own unique ways of looking at the
world and this way of looking uniquely
on the world is a term that is
scientifically called
as cognitive diversity as a visual
analogy consider for example you and me
were looking at the same object a prism
based on how we look at it where we are
looking from we would form different
rules to parse this object we would form
different interpretations of this object
and we will embed different meanings in
this object this is what makes this
cognitively diverse from each other
computationally speaking cognitive
diversity is merely being able to use
different perspectives and breaking
objects in different
opponents using different rules and then
using different heuristics to form new
interpretations so if you’re all
cognitively diverse from each other how
does it affect our creative ability this
brings me to the second most
extraordinary ability of human
intelligence social interaction when we
socially interact with each other it’s
as if we exchange our rules of parsing
the world with each other and in the
process we transform each other’s way of
looking at things together we engage not
just in a dialogue but in a creative
thinking process in the end it doesn’t
even matter whether we understood each
other completely or not as long as we
succeeded in transforming our own ways
of looking at the world so the case that
I’m making is that with just these two
abilities one the ability to be
cognitively diverse and second the
ability to socially interact is what
makes us creative Society however we are
spending less time embracing our diverse
thoughts and ideas we are spending less
time problem-solving with humans and
more times with machines and this
machines are programmed to give us the
exact same answer every single time we
query it so this got me thinking can
machines ever learn to be cognitively
diverse can machines ever learn to give
us different perspectives every single
time we ask it the challenge and making
such algorithms is that creative
thinking is neither strictly logical nor
is it entirely random our thinking is so
ephemeral one moment we might look at
something one way another moment
completely differently it is as if we
dynamically decomposed our thoughts fuse
them embed new meanings and recreate new
thoughts it is as if the units of
thoughts are evanescent in nature and
they serve the purpose of explaining the
meaning only momentarily and then
disappear and reappear to make new
meanings this is because we do apply
rules but we are constantly changing
rules to form new interpretations and
then form new rules again
and making machines that can exhibit
such kind of behavior is computationally
challenging given the current frameworks
with machine learning it’s required
using a different paradigm a
computational framework called shape
grammar conceptualized by George Stinney
a professor at MIT provided a way for me
to address this challenge using shape
grammars we can now build machines that
can compute multiple perspectives
consider an example of these three
triangles if I ask people what do they
see most will say three triangles some
however would see two triangles some
would see four triangles somebody like
me would see nine lines and the
possibilities are endless our human mind
is capable of incorporating all these
different answers machines on the other
hand are programmed algorithmic ly to
give the exact same answer three
triangles but what if our frameworks
were not so static and rigid shape
grammars provides us with a framework to
compute all these multiple possibilities
this is because the machine computes
with new set of rules and new components
every single time the asker it can
either have full partial or no memory of
the previous state
so in short using using shape grammars
we can encompass all these indefinite
possibilities and perspectives using a
same simple mechanism but here is the
incredible part this framework can be
applied not just to compute with shapes
and visual forms in my research I use
the principles of shape grammar and
applied it to creative thinking now
imagine that instead of computing which
shapes the machine is now computing with
abstract thoughts and instead of making
rules of placement of shapes the machine
is now making rules of placement of
abstract thoughts this meaningful
arrangement of thoughts or spatial
relations is nothing but ideas so if
thoughts were shapes not
literally and ideas were spatial
relations then creative thinking can be
computed using shape grammar using this
framework we can now build parser
machines that can break complex ideas
into simpler thoughts we can build
interpreter machines that can combine
simpler thoughts and make complex ideas
or we can make inference machines that
can compute with multiple ideas and
start connecting these thoughts and help
us discover meanings which our human
mind was incapable of discovering in
short I’m saying we can build machines
that can think might not think is
logically written still think and if
these machines can learn to share their
rules of parsing with us they can become
true creative partners instead of being
passive interfaces they can now actively
start interacting with us to augment our
creative abilities my vision for the
future of creativity is not where
machines are more creative than us but
one where machines inspire each one of
us to be more diverse in our thinking
more creative in our thinking with
access to wast knowledge an incredible
computing power machines can now provide
us with different diverse perspectives
as we bring in our diverse experiences
to make the sense of the world so let us
build machines that can help us step
outside our own fixed ways of thinking
so that we can understand the world not
in one way or 13 ways but in definite
ways because like Minsky said if we
understood something one way then he
would not have understood it at all
thank you
[Applause]
Please follow and like us: