so this story begins about 15 years ago
when I started as a graduate student in
a laboratory of dr. Kevin pass/no at the
Ohio State University Kevin was an
interesting guy sat down with him said
Kevin what class is going to take I am a
PhD student an electrical engineering
he said we definitely don’t want you
taking electrical engineering courses
you’re going to get plenty of Electrical
Engineering interacting with me and all
of people around us and your job as a
PhD student is to grow our discipline so
I want you to go out and take advanced
mathematics statistics and actually a
lot of biology go out and take some
evolution courses of course in
behavioral ecology because we want to
grow this thing and we’re not going to
do that by you taking courses that
you’re going to get anyway by doing the
projects here so I thought well you know
that’s kind of interesting and so I’ve
you know in being around people like
Kevin for the last say 15 years or so
I’ve given a lot of thought to what do
we mean by discipline so if you look at
discipline these are some definitions
you can get a branch of knowledge
typically one studied in higher
education or a field of study but when
you walk around the halls of higher
education
you start feeling some of the other
definitions so as one example orderly or
prescribed conduct or pattern of
behavior I often will bump into my
colleagues especially the young ones
I’ll say great you know you you should
meet this other faculty member this
other department they’re working on it’s
interesting problem and I think that
that you you you could contribute and
they tell you know what that that might
be something I can contribute to but not
not right now in my career I need to
stick to what is going on in my
discipline I start thinking so what what
should I do they start looking to other
people in their discipline to say what
are other people in my discipline doing
so our disciplines give us a template
they’re they’re very powerful so we can
ask where do they come from now when I
think about disciplines I think about
things like these Starling flocks so the
beautiful these murmuration ‘s right and
so down at the bottom it says we’re
disciplines come from
and in these Starling flocks when we
look at these we might you know wonder
do these do these
formations do these flocks actually have
any function but in reality these flocks
themselves a biologist will tell you
that the actual shape of these
macroscopic structures don’t have any
function they’re actually the result of
each Starling playing it safe each
Starling is trying to minimize its risk
of predation so it gets as close to the
other starlings around us and you get
these beautiful structures that’s what I
think of our disciplines sometimes is
that we get closer and closer to those
around us to play it safe and we form
these structures that we worship but in
the end it’s not clear that they’re
functional fact sometimes when I think
about that that faculty friend that I
talked about who I said you know you
should go meet up with this other one
well they’ll tell me is no you know I’ll
do that after I get tenure discipline
punishment inflicted by by way of
correction in training discipline to
bring a state of order and obedience by
training and control discipline to
punish or penalize in order to train and
control to correct to chastise again in
that story I’ll tell these these guys
you know you should meet up you know
faculty a and faculty B you should come
together and I’ll say if I did that
there’s no way I could ever get tenure
those are things I save after I’ve
gotten tenure so discipline it’s it’s
very controlling and maybe it’s a good
thing but we have to ask what costs do
we pay when we take these disciplines
for granted and that brings me back to
another story that started back when I
was a grad student so this is a
beautiful picture taken by Alex Wilde of
a nigma Newman wasp but parasitoid wasp
and when I was with Kevin and grad
school Kevin said when you look at that
don’t think of that as an object for
entomology or an object for biology when
you look at that you’ve got to realize
that that fits into a set of things that
go right along with this autonomous
helicopter or this autonomous underwater
exploration vehicle or this autonomous
vacuum cleaner they all are making
decisions and we need to view them by
what they’re doing and not who studies
them you’d say well what sort of
decisions is this wasp making well
in the 80s some behavioral ecologist
looked at wasps like these and they said
you know it’s interesting these wasps
have a finite number of eggs and they
have to lay them in a finite number of
places and they can’t lay the same
amount of eggs everywhere they have to
decide how many eggs to lay in each
place so they picked up operations
research textbooks off the engineering
shelves and built these optimization
models the details don’t matter but
these mathematical models and they
accurately predicted where the wasp was
laying those eggs but wasps aren’t doing
calculus in their little heads so then
they had to go forward and say how can
we build simple behavioral algorithms
that when combined with the environment
end up producing outcomes which follow
these optimality solutions and and when
Kevin and I read that we said you know
this is great we could use the same
tools for engineering the problem is
that the functions the behavioral
ecologist use weren’t quite general
enough for our engineering context so
rather than switching to a new framework
we expanded their framework so we could
handle the whole family of autonomous
decision-making and we published that in
a robotics journal expecting that
roboticists would be interested and most
interested in it but it actually got
picked up by a wireless communication
the set of people working on on multi
radio networks we were trying to figure
out new ways to allocate energy to
radios to provide the same bandwidth at
a lower cost and using our methods we
developed from this wasp they were able
to provide the same bandwidth for lower
energy than off-the-shelf methods now
they put foraging in the title of the
paper that they published here and they
knew that foraging theory had something
to do with it but I’m sure that they had
no idea that a wasp was behind it all
now you might ask why didn’t I talk more
about the wasp in that paper because it
was valuable for me to know about the
wasp so it should be valuable for them
to know about the wasp well the thing is
that wasp is too big for that little box
that little paper there if I would have
gone on and on and on about these
parasitoid wasps and that paper it would
have never gotten published in a
robotics Journal the editor would say
save that for another community our
community is not interested in that
keep it to the robotics split this into
two papers because there’s no reason for
one community to read papers from that
other
unity so we kept it out but obviously it
has some value at least we think it has
some value I think this one has slightly
other costs so here’s another one a
definition that I came up with it
provides an opportunity to reinvent the
wheel again and again and again it’s
another example the math on here doesn’t
quite matter but this is a snippet from
a paper written by a Nobel Memorial
Prize winner in economics William sharp
back in the papers from 1966 he won the
Nobel in 1990 and this is a very
influential paper that tells financial
planners how to allocate money to risky
stocks so they could maximize their
performance at the lowest aggregate risk
and it involves this thing called the
Sharpe ratio which shows up in the
middle of that equation now later on
some behavioral ecologist like the ones
who work with that wasp started saying
how do animals allocate their time in
different ways that also have risk and
they wrote a paper that has actually the
exact same ratio in it now you might
think that they read the financial
literature just borrowed it directly
from him you know hurrah for
interdisciplinarity but unfortunately
they derive this all by themselves
totally ignorant of this work over here
so the Sharpe ratio even though it’s
named after a Nobel Memorial Prize
winner is was totally gone unnoticed
when I presented this in my Master’s
defense a behavioral ecologist in the
audience said Eureka finally have
something to call it when I teach this
stuff but I mean seriously if you look
at the gap here 1966 to 1980 – that’s
almost 20 years where these communities
could be working together but they
didn’t because what do people in finance
care about insects and what do people in
insects care about finance what do I
think discipline should be this one to
me is a branch of instruction or
learning we teach our disciplines but we
practice our problem-solving I run a
small lab here at ASU I call the science
and engineering of autonomous
decision-making systems people sometimes
ask me what do you do in your lab and I
give them this picture where I have this
I say well we link together all these
different disciplines some computers
and cognitive psychology and it’s a nice
picture but in the end all we really
care about is whatever tools allow us to
study autonomous decision-making in
natural and artificial systems and I
should emphasize natural and artificial
systems both equally together not one in
service of the other now when I first
came to ASU they’re pretty excited about
this when they hired me and I was pretty
excited about the interdisciplinarity at
ASU but when I started talking to sort
of the middle management of ASU allocate
resources their more traditional
engineers and they started asking me why
do you need a laboratory
can’t you just collaborate can’t you
just go to all these people you know
here at ASU and have them run your
experiments for you again they brought
up the tenure thing there’s no way
you’re going to get tenure
if you’ve running experiments in a lab
you’re an engineer what are you doing
run while you’re asking to run put ants
in your laboratory but that doesn’t
sound like collaboration to me that
sounds like parasitism so when I
interact with my collaborators it’s more
about learning how they think different
directions on the same problem we are
united by the problem not about the
resources we need to get let me give an
example of that so here’s another Nobel
Prize winner this time in Physiology and
medicine Niko Tinbergen he’s one of the
fathers of behavioral ecology now
behavioral ecologist are interesting
when they talk about communication they
group things into signals and cues
signals are the messages that we
explicitly send from one person to
another for the benefit of both and cues
are sort of things that we give away
even though we don’t mean to and
behavioral ecology will tell you that
every signal evolved from a cue that was
once co-opted well I had a graduate
student in computer science who’s very
skilled at machine learning and he said
that’s interesting what if I can take a
bunch of robots in a string together
that you’ll see here and restrict them
from sending any messages to each other
all they can do is watch the robot right
in front of them and I’ll have the lead
robot we’ll hit an object and the goal
is for all of the rest to eventually
surround the object what he found out is
he could actually use the slight Wiggles
of the robot in the sort of the
second-to-last robot and the final robot
could use those Wiggles to infer the
positions of
every other robot and thus be able to
coordinate its motions based on what
happened to the lead robot with
absolutely no communication that just
came from thinking about the problem
like a behavioral ecologist we also take
examples from engineering and apply them
directly to ants this is another one of
my students who actually works in the
lab with ant colonies using theories
that come out of thinking about
intelligent lighting systems we go
further and further we have information
processing where we have a team of a
multidisciplinary team where we actually
have a swarm robot assisted of the
experimental work with the ant and
somebody who studies the way planets
break apart and come back together doing
all of the modeling work with the ants
we have another recently started project
we’ve hired a tropical ecologist to come
and do work with these beautiful army
ants all in the service of understanding
how information will be passed around
the internet of things and this is a
project where although we had all the
skill sets to come together to put this
together the military actually came up
with this idea they said we don’t see
anything really valuable in a lot of the
way the internet the things is being
organized with this particular problem
in mind we think insects might have
something to say can you do something
for us so looking ahead we have a lot of
these other projects where we’re just
basically I’m not going to go through
all these but the idea here is we’re
grouping people together by problems
they’re interested in and not resources
that they have so I guess the message
for you and thinking about innovation is
I’m hoping that as you move forward and
try to be innovative be undisciplined
what do you mean by that remember that
your discipline is what you teach your
discipline is not what you learn and it
certainly shouldn’t be defining what you
do thank you [Applause]