Complexity | Eric Lumer | TEDxLakeComo

This talk was given at a TEDx event using the TED conference format but independently organized by a local community.

Learn more at http://ted.com/tedx

Tonight I want to explore the relationship
between two complex phenomena.

Complexity is often a dynamic feature,
rather than a static one.

It's a process.

As such, we heard it earlier on,

it’s very difficult to represent visually,

because it develops over time,

in many spatial dimensions or networks.

These two complex processes

are related to the systems and networks
allowing communication on the Internet,

and to the neural bases of cognition.

In other words, how our brain works.

I was inspired to find a relationship
between these two topics,

which I devoted most
of my professional life to,

by a talk by Kevin Kelly,
executive editor of Wired magazine,

given at TEDGlobal two years ago.

Kelly was discussing
the first 5,000 days of the Web,

wondering about one question:

what can we expect
in the next 5,000 days?

He highlighted the fact
that in a very short time, a few years,

the Web has become
an extremely complex system.

First, in terms of its structure.

For example, it contains
55 trillion hypertext links between pages,

a number so big,
it means almost nothing to us.

Second, in terms of the data
that it’s continuously conveying:

100 billion clicks per day,

even before the boom of social networks

and the amount of information
they are constantly exchanging.

So much so that these figures

are beginning to be comparable
to those of a human brain.

If we look at the number of components

in this huge global machine

and compare it

with the neurons in the human brain,

we realize that the number
of connections in the Web

is approaching the same order of magnitude

as that of synapses in our brains,
at least as far as neural circuits go.

This brought Kelly
to speculate on the fact that

on the one hand, this system is becoming

a way of constantly accessing information

for all human beings,

but on the other hand,

we're turning into collectors
of information,

endless sensors of this giant machine.

We'll be going around with
these "sixth sense" devices you've seen,

inputting information
and sending messages over the Web.

In essence, telling it
what’s happening around the world.

There'll be hundreds of millions of people
doing it, soon maybe even billions.

The same applies to artificial sensors,

which continuously provide the Web
with data on environmental conditions.

At the same time,

these same people and devices

will become actuators of the Web,

allowing it

to interact with the world
and actively change its course.

Indeed, Kelly speculated on the fact

that we are actually
developing a big machine,

what he calls the ONE, an ethereal entity,

which might start to display
some features of human-like consciousness.

Having studied both topics,
I got interested in this question

and I’m afraid I must disappoint you.

The Internet is a packed house
but no one lives in it.

There’s nobody in there
looking out at what’s happening.

It’s as conscious as my shoes.

However, what I’d like
to illustrate tonight

is the fact that, gradually,

it's acquiring a series
of traits and capabilities

that enable it to host

ever more coordinate collective processes

which have some of the features
of human consciousness:

the ability to anticipate
what's happening,

the ability to reason and take decisions,

even the ability to think about itself,

to be self-referential.

These collective processes,

as there will probably be more than one,

will have a deep impact

not only on the way our societies work,

but also on our very notion
of social consciousness,

which is actually being redefined.

Let’s have a look
at what makes up our brain

and at some of the ingredients which,
while not sufficient per se,

are essential for the development
of consciousness.

First of all, the ability to learn:

that is, the ability to interact
with the external world,

and to absorb features, correlations,

and representations of the external world,

as well as the way we interact with it.

The most concrete
manifestation of this ability

is the selection and modulation
of connections among neurons.

These synapses are responsible
for the propagation

of electrochemical pulses
from one neuron to another

and for the creation
of these systems in our brain

according to our senses
and our locomotor systems.

Anyway, this plasticity,

this ability to modulate connections
between neurons in the brain,

is not a passive system
left to its own devices,

limited to absorbing relationships
and correlations from the external world.

It’s a learning system
which is driven by values,

by systems that are
actually based on values.

Such value systems are located
in the inner kernels of our brain.

These neural circuits lie
in the archaic regions of the brain,

the same that we find in all mammals,

and deeply influence

the whole of the cerebral cortex.

They have a tendency to use data
to create a raw interpretation of reality,

contrarily to what the cortex does.

It’s a basic interpretation
of what’s happening around us,

an instinctive first impression

that attributes a value
to what’s going on.

The cerebral cortex
tends to build representations

integrating them with our conscience.

Instead, these neural circuits
assign them a value.

For instance,

their activation can prompt us
to satisfy our basic needs

such as eating, sleeping or reproducing.

The way these circuits work

is genetically inherited.

In turn, the cortex is capable
of influencing such circuits

and modifying them

to create systems
of so-called acquired values,

which aren't inherited

and are the cause of certain behaviours

that were also discussed here tonight.

For instance, the development of ethics,

shared concepts

which are largey acquired
through social interaction.

The third element

is one of the key neural bases
of consciousness.

It's the great ability to integrate

all modules and activity
components within the brain.

Why is this?

Because consciousness is unitary.

It has to be, as it is strongly related
to our ability to plan

and organize our interactions
with the external world.

In addition, we have only one body.

Imagine what would happen

if there were two conscious entities
within the same body,

and one decides to turn left,
but the other goes right;

we would have trouble walking!

Yet, in some extreme cases,

this can happen.

Our consciousness
can be split experimentally,

as we’ve observed in split-brain persons.

Roger Sperry was one of the first
to do research in the field,

winning a Nobel prize for it
about 30 years ago.

What happens to split-brain people?

They have their corpus callosum severed.

This disrupts the major connection
between their two hemispheres,

which carries neural signals

between the two halves of the brain,

which allows this integration
of information between them.

After this corpus callosum is severed,

which usually happens in treatment
of severe epilepsy,

the patient appears to be
a single human being

for most of the time.

Actually, in some specific conditions

which can be created,

we can observe the presence
of two reasoning entities.

Basically, the patient has a barrier

positioned between the two visual fields.

As you probably know,

the right visual field
maps onto the left hemisphere

and vice versa.

Then they're given some sight
or voice-related instructions,

"According to the word
you’ll see in front of you,

you shall either say it aloud

or take the corresponding object."

The left half of the brain,
which is linked to the right visual field,

sees the word "RING".

Instead, the right hemisphere,

which is responsible
for language and speech,

thinks: "I see the word RING,"

causing the patient to say: "Ring".

Simultaneously,

the right half of the brain
sees the word "KEY"

but can't speak it, not having
any concept of verbalization,

so the patient takes the key in his hand.

We have an individual

who takes the key
and says "Ring" at the same time.

This demonstrates how integration

is crucial to the development
of cognizant behaviour,

and how, in some cases,
it can be interrupted.

Besides integration,
a great ability to remap is needed.

That is, to constantly reorganize

what’s happening within the brain.

Basically, remapping means
taking what is received

from the external world,

in this case, through the eyes

that detect the light reflected by objects

onto the retina,

and combining it

with our relevant value systems.

This is continuously carried out

on past perceptions,

up to a few seconds in the past,

and constantly updated as time passes.

We remember and as we move on,
this changes our perspective.

It’s as if the back of the brain,

more concerned with raw signals processing

and feelings from the outside,

were constantly telling the front
what’s going on around it.

The front then tells the back

how it must interpret those data,
according to what it remembers.

That’s what Edelman calls
"remembered present",

a continuous remapping, a dynamic process

that seems to underlie
the most basic form of consciousness.

We're always speaking in the third person.

The last ingredient is narration ability.

This is the ability
to remap this representation

of what’s going on and what I'm doing.

Or rather, what’s going on
and what the brain is doing

in a self-centered perspective,

according to an intuition of the self

that’s determined mainly
by the fact that we have one body

and that everything our brain perceives

is connected to its sensors and actuators,

which belong to our body.

Hence, this narration,

that is, this remapping,
in a self-centered perspective,

happens over time.

In this system, actually there is
no entity watching what’s going on.

There’s just a process,

chatting with itself
about the meaning of the situations

in which it is.

From here, it can start building
abstract and symbolic representations

ultimately creating
both symbolic concepts of situations

and a self-concept,

a definition of its own identity.

Now you should have a basic idea,

not necessarily of what this means,
because it’s a subjective experience,

but rather of what physical ingredients

are needed for this process to take place.

And in particular,
of the activities and relationship

between the rest of the brain

and the prefrontal cortex.

If we consider various extreme cases

when conscience is dramatically inhibited,

if not suppressed,

such as coma, vegetative state,

the more common deep sleep,

or also general anaesthesia,

we see that, in all cases,

their common denominator is hypoactivity,

that is, reduced activity
of the prefrontal regions

compared to the pattern of activity
that can be observed in normal people

who are able to verbally give proof
of their consciousness.

Finally, it is no coincidence
that consciousness and self-awareness

seem to arise almost exclusively

in social animals.

A key factor in their development

and the discovery of their self-identity

is the ability to distinguish
between self and non-self.

This learning process

is greatly stimulated

by interaction with other members
of the same species.

There’s a very restricted club of animals,

which includes great apes,
humans in particular,

and was recently joined also by elephants,

that are able to pass
the so-called mirror test.

When in front of a mirror,

they understand that
what they see is a reflection

and not another animal.

Also, if we paint a cross
on the forehead of an elephant

while it’s sleeping,
once in front of the mirror,

it will try to erase
that cross with its trunk.

Other animals will simply try
to interact with the mirror.

So, self-awareness is
an individual perception.

Yet it’s a social sense,
which arises in society.

But now, let’s go back to the Internet.

The Internet has developed
its infrastructures,

its ways to obtain information

and the ability to influence
the external world

on a massive scale.

This is particularly true
for social networks.

These are constantly reorganizing

and operating an actual rewiring

of the social graph
of connections between people,

affecting the way information spreads

among the nodes of the network,
which ultimately represent ourselves.

As for the values, these were designed

and embedded into the system
by its founding fathers.

Such as end-to-end equivalence,

which means that there are
no special, differentiated nodes

and all the processing
is done by the end nodes:

the one that sends

and the one that receives signals
over the network.

It’s either an open system
or a federal one.

On top of these principles,
acquired values appear,

based on the constant feedbacks

that users and participants
to this network

give about any bit of information
that is uploaded on the Web.

They do it through systems like those
that you see on any website these days,

which let you share a piece of content,

rate it or leave a comment.

These are nothing more
than value judgements

that collectively contribute

to the organization of content

and information circulating on the Web.

Digging deeper, we discover

that there are also integration processes,

which can be monitored and tracked.

For example, the so-called
information cascading processes,

according to which the probability

that I share a piece of information

is directly proportional

to the number of people
that have already done so.

Such cascading processes,

which can cause
some specific bits of content

to rapidly go viral at some point in time,

are the basis of phenomena

like the one that occurred
during the elections in Iran.

Here, the popular revolt
first broke out on the Internet.

Why?

Because there was no general perception
of the level of dissent.

Only thanks to this way of gathering data,

more and more people started to realize

how widespread dissent was

and this ended up in massive protests.

Finally, we are witnessing
the development of an ability

to classif,y in greater
and greater detail, any information

that is uploaded or conveyed by the Web.

This is thanks to these systems
of relationships among people,

virtual and real objects,

which are constantly reclassifying
everything on the Internet.

Yet, we are still missing
the narrating self.

This system is still trapped
in the present.

There are no processes
to analyze data over time,

which would enable the Web
to take decisions and predict the future.

Yet, some people are working on it.

These are people that are used

to deal with and organize
massive amounts of data.

So, to summarise, I don’t think
the Internet is conscious.

It’s more likely to develop
some forms of precognition

or collective organization,

which will be able to predict

and influence the behaviour
of large sections of the population.

There will probably exist more than one.

And they will be capable
of having a dramatic impact

on our civilization.

I believe that any piece of information
created across the world

will end up on the Web in some way.

This is getting increasingly smarter.

It is getting less limited
by data and time

in its processing of information.

There won’t be just one
but rather several collective processes.

And they will redefine

our concept of collective conscience

and what we mean by it.

I guess some of these processes
are already taking place.

Thank you for your attention.

(Applause)