Robots, Collective Intelligence, and Participatory Governance: Nikolaos Mavridis at TEDxCarthage

After a short history of intelligence on the planet Earth, starting from the first organisms and arriving at intelligent robots, the talk provides examples of various kinds of robots that interact with humans. Then, it touches upon the future of hybrid human-machine collective intelligence, and after discussing new technologies regarding participatory governance, arrives at a vision for the green smart cities of the future.

Imagine a world

where robots are part
of your everyday life,

and not only as helpers,
but also as companions and friends.

And where the minds of such robots
are not just building the big,

technological centers of the world
like Silicon Valley,

but they are built even
in a small oasis city

in the middle of the Arabian Desert,

such as Al Ain.

Furthermore,

imagine a world

in which we can all participate
in wider intelligent entities,

that are made up of potentially
thousands of humans and machines,

even with ten seconds
of our time every day,

and whose capabilities might
far surpass the current limits

of both human
as well as artificial intelligence.

And finally,

imagine a world

where we can harness

the collective mind and intellect
of the citizens of our nations

in order to be able to bring forth

effective participatory governance
for the smart green cities of the future.

These three have been my dreams
and goals in the last ten years.

And they have partially
produced some fruit:

such as Ibn Sina, the world's first
Arabic speaking android robot.

In the words of the BBC,
as recorded in my lab:

(Video) But unlike his older counterpart,
this Ibn Sina is altogether

more high tech.

As well as being
fluent in English and Arabic,

he can hold a conversation with humans.
He looks for information about us

on the Internet,
and uses that to chat with us.

He'll also remember it
for the next time you meet.

Robot voice: Saeed - Happy.

Narrator: But this lab isn't
in Japan, China, or the United States;

it is in Al Ain, just an hour's drive
from the UAE capital Abu Dhabi.

(Applause)

Nikolaos Mavridis:
I would like to ask you

to embark with me upon a journey,

that will have three parts:
first, I will start with

a very short history of intelligence
on the planet Earth, to arrive

to the moment where we actually
have robots interacting with humans.

Then, I will talk about
individual intelligence,

and how this scales up
to collective intelligence.

And in the end, I will go
to the most crucial question, I think,

for today: how can we harness
the collective mind of our citizens

in order to bring forth
effective participatory governance?

Let us start!

In the beginning, the Earth

was full of lifeless matter.

And then,

the first organisms appeared,

And later,

the first animals

that could actually perceive
the world through their sensors,

their eyes and their ears, and then

mediated through their
natural intelligence,

they could act upon the world

through their muscles.

Later,

humans appeared,

and with them,

the first man-made objects,

such as tools,

and much later,

a very special kind
of man-made objects:

machines, that could
run along on their own.

And it wasn't until very recently,

that we started getting
the first intelligent machines,

such as computers.

But ask yourselves:

how much has our life changed,

since the introduction
of intelligent machines?

Just think:

in the last week,

how much time have you spent

in direct interaction, one-to-one
with another person?

Versus how much time have you spent

in electronic mediated interaction,

through your cell phone,
or through your messaging system?

Versus, more importantly,
how much time have you spent,

interacting with a machine,
and not a human?

With a website,
with a computer game, with an ATM,

and in the future, with robots?

The average American is spending

two hours of his day online,

and a big part of this time has to do

with interaction with machines,
not with humans.

So we are spending more
and more time with them,

and slowly,

we are starting to relate to them.

And then,

came robots.

And robots are very different
from other intelligent machines;

they don't need to have
somebody at a keyboard

in order to feed them

with information from the world,

as you need for a computer;

they have their own sensors,

they can perceive the world through
their cameras and their vision systems,

through their sonars, through their
auditory sensors and so on.

And most importantly they can also
directly act upon the world;

they have their own muscles

which are their motors, in this case.

And the connection between
the sensors and the motors

happens through their
artificial intelligence.

For example,

when it comes to chess,

the Deep Blue computer of IBM

was able to win over Kasparov,

the World Champion of chess.

But still, in many, many other things,

human intelligence is far superior
to machine intelligence;

take for example, the motor
abilities of a 5 year-old girl,

if you get the best biped humanoid robot,

like Asimo that you saw,

it is much worse when it comes
to what it can do.

Take also the abilities of young children

to be able to continuously learn

in an exploratory manner.

Again, the best machine learning program

out there cannot even think about
doing something like that.

There are unsurmountable limits, though,
to what a single intelligent entity,

either it being human or being machine,
can do on its own.

And this is why, participating in groups

is not only very, very pleasant
as an activity,

but it can enable us to do things
that we cannot even think of alone.

But the intermediate stage

between the single individual
and the group

is the pair.

In this case, a pair of a robot
and a human

interacting with one another.

And most of the work
that I have done so far in my lab

actually has to do
with exactly this setting.

And let me share with you
three interesting examples of that:

first, the robot Ripley,
which has the form of a helping hand.

You can give it commands in language,

it helps you with things
that you want to do on a table

and it can learn the meaning
of words through examples.

Let us see it in action!

(Video) Man: Where is the red one?

Robot voice: At the top.

Man: Look at the center!

Robot voice: OK.

Man: Where is the red one?

NM: So up there you can see
what the camera sees,

and this is the model
of the world that it makes.

(Video) Man: What color are the objects?

Robot voice:
One is red, and the other is green.

NM: It cannot see the red object anymore,
but it still remembers it.

Man: Imagine a blue object on the left.

NM: Now we are telling it
that there is an object on the left,

so it's imagining an object

and it waits to see it
when it looks on the left.

(Video) Man: Where is the blue one?

Robot voice: Most probably at the left.

Man: How big is the blue one?

Robot voice: I have no idea.

Man: Look at the left!

NM: And now he will see
what he was imagining before

and verify that indeed
what I was telling him was true,

that there is a blue object,
and it will know how big it is.

Man: How big is the blue one?

Robot voice: Small.

NM: And the robot can also remember.

Man: How big was the blue one
when your head started moving?

Robot voice: I had no idea.

Man: Where was the red one
when the green one appeared?

Robot voice: Most probably at the top,
but maybe not.

Man: Thank you!

Robot voice: You're welcome.

(Applause)

NM: So you see how, slowly,
these machines are learning

how to bring part of the world
within their own mental realm, their head.

And then they're learning how to connect

all of these things with language
and help us,

and even remember
in some cases, and imagine.

The second example
that I want to share with you

is from the Facebots robots.

And here my purpose was to build
a robot that can interact with you

not only once,
but it can keep your interest

so you can speak again with him

after one day, after one week,
many times within a month

or three months.

And for that reason it's trying to create

shared memories with you,

so it remembers the things
that you did together,

and shared friends,

so it can spread gossip about the friends

and have an interesting discussion.

At the same time it can meet with
people physically or through Facebook

and also it can get information from
Facebook that is relevant to them.

Let us see that too!

(Video) Robot voice:
Are you Layli Tutunzin?

Woman: Yes!

Layli Tutunzin, are you doing well today?
Woman: Yes!

Robot voice: Well, let me tell you
some updated news

from the world of technology.

NM: Here it is giving customised news.

Robot voice: That's great!

NM: It has recognised the face
of the person.

Robot voice: Layli, do you know
that our friend, Panos Tulli,

has posted a new photo
in the photo album?

Woman: No.

Robot voice: I am sending you
an inbox message about this.

NM: Talking about common friends.

RV: One of our common friends,
Panos Tulli, is online.

Would you like to send
a message to our friend?

NM: It can connect people, one of which
is online, the other physically there,

even get people to know one another.

RV: I remember when we talked
this morning and I had asked you

"Are you doing well today?"

And you were doing well

and we talked a little about Chandan Data.

Layli, it was nice speaking to you!
Goodbye for now.

(Applause)

And then, the third example

after Sarah the Facebot

is Ibn Sina, which you very
briefly saw in the beginning.

And here the purpose was
to create a shopping mall robot,

or a museum robot,
or even an Interactive theatre actor.

And this robot actually spoke with
thousands of people as you shall see.

A lot of degrees of freedom in the face,
it can have facial expressions.

(Video) Man: Salam Aleikum! Marhaba.

NM: A simple dialogue system in Arabic.

It even does the greeting

of the Khaliji greeting. Haha!

Robot: Hazin: sadness.

NM: And it can take facial expressions
corresponding to emotions.

Sadness.

Robot: Elpostir: disbelief.

NM: Disbelief.

Robot: Motahamas: Excited.

NM: And excitement!

(Laughter)

As you can imagine
this got a lot of publicity,

so just a small sample follows:

This is Agence France Press.

(Video: Speaking in French)

Dubai TV.

Thirty countries around the world.

Netherlands.

And then the robot also went
to shopping malls and exhibitions

and spoke with around 3000 people.

And it also flew on Emirates Airlines
to Saudi Arabia,

invited by the Minister of Education
of Saudi.

And here he is very happy.

And in the end the letter that he wrote

to Sheikh Nahayan
bin Mubarak of the Emirates.

(Applause)

So all of these three examples
that I showed to you

were examples of interaction
between pairs:

a human and a robot,

speaking with one another.

The next stage is group interaction,

and in this case we have a group
of two robots and two humans

from the NASA Robonaut Project.

And after that, what we reach
is what one could call

in Layman's terms "superorganisms"

that could have thousands
of humans and machines,

connected tightly to one another,
and sharing their eyes and hands,

and brains, if you want.

But what do I mean exactly by that?

Let me give you an example.

These are ten red balloons;

they are about I think
five meters in diameter,

and they were placed in random locations
around the United States in 2009

as part of a DARPA research program:

this was the network challenge,
as they called it.

And the goal of the teams
that were taking part in the competition

for these balloons
which were University teams

was to find a way to find them.

And they could use whatever they want:

they could use humans,
they could use satellites,

they could use machines.

So the team that
was actually able to win

which came from the MIT Media Lab,
where I did my PhD too,

used - what do you think?

Social networks.

So they used social networks in order
to get to recruit thousands of people

who pretty much spent ten seconds
of their time lending their eyes

and their brain to the system
to look around

and check if they see a balloon.

And then, through the thousands
of eyes that contributed to the system

they were able to find them
in a very, very [short] time.

Of course you can extend this.

You can get people give part of their
brain abilities

as services to the system.

They can be translating text,

they can be planning,

they can be doing operations.

And if you bring all of this together,

then you get something that
is much more similar, if you want,

to this dance of the reincarnation
of the Buddha that they have in China.

So enough with interactive robots
and the human-robot cloud.

Now let me go
to the third part of the talk,

and talk to you about the answer

to the following question:

how can we harness
the collective mind of citizens

in order to be able to bring forth
effective participatory governance?

Both in Ancient Athens as well as
in Islam and many other cultures

one of the prime elements of politics

is "consultation" or "Shura"

as it's usually called.

The big issue though with Shura
is that it works at the level of a village

but how can you extend this
to the multi-million-people nations

that we have today?

This is a big problem,

but this is exactly
the opportunity in terms of technology

for the first time in the
history of humanity to try to do it.

And a prime example of this,

is what his Excellency

Sheikh Mohammed bin Rashid
has done in Dubai,

in which through
the UAE Brainstorm program,

he brought about ten thousand
of his citizens to contribute

to the reformation of the
education program of the Emirates

by sending messages through Twitter
and through the Internet.

And there are many cases
in which electronic consultation

can really help bring forth new ideas

and can really help people
participate in what is happening.

There are cases though where
it is important to, if you want,

get the physical element come in there.

That's why, as part of my involvement

in “Recreate Greece", which is
a citizens movement in Greece,

I planned and organised hybrid
electronic and physical workshops

for consultation, in which again
we had many citizens of Greece

coming together and brainstorming

in order to plan for the
Renaissance of Greece after the crisis.

And the important thing
about the physical element

which you cannot find too easily
in the electronic

is the fact that you can have

real friendships
taking place in this meeting,

and also the citizens that get involved
later start group action together.

So you actually get volunteer groups too
that really do things and not just talk.

And there are many other technologies

that have become available right now

to be able to enhance governance,

that are based on this collective
mind of the people, if you want.

There is liquid representation,

there is participatory budgeting,

there are online petitions,

and much more that you can easily find

if you see for example the World
Forum of Democracy in Strasbourg.

Also, there are many new technologies

regarding being able
to analyse public opinion.

In a project that we started in my lab
recently, we actually fetched

150.000 tweets that were related
to the ObamaCare website

and how people reacted to it.

And we were able to find out

who the main influencers were,
what the spheres of influence were,

what the communication dynamics were,

how the sentiment of the people

was coming up and down over time,

and how this whole thing was related
to what was happening out there

in terms of government announcements
and the real project progress.

And you were even able to actually get
predictions from what is happening,

through this analysis of public opinion.

Thus, in the green smart
cities of the future,

it's not just the infrastructure
that will be intelligent

and the buildings,
and environmentally friendly.

One of the most important elements
we think in that respect is to be able

to harness both the skills and the mind
and the opinions of all of the citizens

in order to help governance,

and in order to help them
make their dreams come true.

Thus, we have gone a long way.

I started by talking about lifeless matter
and went all the way

to robots interacting with humans.

Then we talked about
individual intelligence

and collective intelligence.

And finally, we asked about
the collective mind of the citizens

and how it can help
in the new cities of the future.

Let me thus just close
with a vision of the future.

Imagine a world in which humans, robots,

and all other intelligent
and natural entities

that will make up the complex biological
and artificial ecosystem of tomorrow,

will co-exist in peace,
harmony, and mutual benefit,

and in which through
our participation in larger entities

we will able to reach a much more
deep collective self-actualization.

And finally, imagine a world,

in which through
the collective mind of the citizens,

we will be able not only to have
effective participatory governance,

but most importantly,
we will help preserve and enjoy

our most beautiful planet,

its biosphere,

and Mother Earth.

(Applause)