Networking Faster Than Light: Alex Wissner-Gross at TEDxBeaconStreet

The performance of a wide variety of globally-distributed online activities is increasingly limited by the finite speed of light. In this talk, Dr. Wissner-Gross introduces the technology for partially mitigating the impact of this limitation on the coordination of geographically distributed activities.

We live in a physical world.

We also live in a physical world

that's in a process of
merging with the digital world.

As a consequence,

increasingly, aspect of the physical world

such as, say,

the fact that it's
geographically distributed,

are starting to impact
our digital experiences.

This is a worldwide property,

and it's due to the fact

that we live all across
the surface of a planet

that has a finite diameter.

So, for example, in virtual worlds,

if you have one party
on one side of the planet,

and one party
on another side of the planet,

that are interacting via a virtual world,

such as Second Life,

the experience delays due to the fact

that light takes a finite amount of time

to travel around the Earth's surface

in order to connect them.

And this problem isn't in any way

specific to virtual worlds
or entertainment.

It's a problem that's keenly felt
in industrial sectors,

in financial sectors.

Financial sector in particular,

is very well incentified to make sure

that the time delays
for transmitting information

between financial exchanges is minimised.

One more example,

telerobotics and telesurgery.

As we move towards an era,

where a physician
in one location on Earth

can perform surgery on a patient

on the opposite end of the Earth,

the delays involved in
manipulating remote machines

become increasingly essential
to performance.

And so how if we try
to combat this thus far?

This is a map of submarine cables

strewn all around the Earth's
oceans and on land.

We are literally
wiring up our planet surface

in order to efficiently
allow information to flow

from any point on the Earth's surface

to any other point
on the Earth's surface.

You might imagine

we are relatively close to

solving this problem
of information delays.

But of course,

as with all physical properties,

there are limits.

Here's one very important one.

What you're seeing here are 2 maps

of some of the state-of-the-art

Internet connections
connecting locations.

On the left,

a New York to Chicago connection.

On the right,

New York to London.

Interestingly,

if you look at
the amount of time it takes

to send information back and forth,

through these pipes,

and you compare it

with the theoretical physical limit

to how fast you could
send information round trip

using light through optical fiber,

you'll notice

that we're literally approaching

the physical limits allowed to us

for sending information around the Earth

between these important cities.

This is the problem
for the reasons I've mentioned

that's only going to become
more exacerbated with time.

Why is this a problem?

Well, the way we've architected

many of our global transactions
right now on the Internet,

and our networks in general,

requires round trip transactions.

So, you have location A,

and location B,

and you want to coordinate processes

between these two locations,

they have to send
information to each other

about their current state,

and wait for the other party

to react to that new information,

and send back an execution signal.

In other words,

for most transactions today,

you're stuck with round trip delays

on the Internet.

Now, biology,

as has historically been the case,

has come up with
a solution to this problem.

So, if you as a human,

touch a hot surface,

you'll note,
that your hand pulls itself back

before you feel the pain
in your mind.

And the reason is that the pain signal

isn't travelling the entire distance

from your finger up to your brain,

and then back to your motor neurons,

to pull your hand back.

Instead, it's just travelling
to an intermediate location,

your spinal column,

where it's immediately recognized

as actionable information,

and signals are immediately
turned around

for a more rapid response.

It's called a reflex arc.

So, biology is handing us the solution.

We can do reflex arcs
for global networking.

So rather than doing

round trip transmission of information,

let's instead position
server infrastructure

at well selected intermediate locations

between two different
geographic locations

that need to be coordinated.

So, this is a problem

that I've been thinking about

for a number of years:

how best to position
these intermediate locations

in order to literally
get around the speed of light

by not being forced to
wait for this round trip transmission?

And recently,

I identified and published
the optimal solution.

So, what you're seeing here
is an equation

that describes
the theoretical optimal solution

for where to place yourself

in between two different locations

in order to optimally
coordinate processes

happening at either end.

And so, just for fun,

I've taken this equation,

and applied it to a financial task.

So, what you're looking at here,

is literally a treasure map.

The large red dots represent

the world's largest stock exchanges,

and the small blue dots here

represent the calculated
optimal intermediate locations

for coordinating, trading,

on pairs of these exchanges.

And you'll note,

that many of these
optimal intermediate locations

are in network sparse areas,

and on oceans.

And so, this presents us with
what may perhaps be

the first excuse to
literally boil the oceans

using computation.

Now, how do we go about to
deploying this new infrastructure?

So, I've shown here
3 different modalities

for deploying these new servers

that would be responsible
for creating global reflex arcs,

in order to help us get around
speed of light limitations.

So, on the left,

you see microwave towers,

we see balloons in the middle,

and we see microwave buoys,

microwave relay buoys, to the right.

These are just 3 modalities
by which one can envisage,

getting around the speed of light

due to the Earth's finite size.

And so, I think, stepping back,

we're faced with an opportunity
at this point in time,

to not just bump up against

the finite speed of light,

as a pretence to
telecommunications on Earth,

but to use this as an opportunity

to create lots of
new infrastructure in locations

that previously had none,
and in particular,

I'm especially fond of this analogy,

the Silk Road, or the Silk Route,

the trade of silk was responsible,

in the first few centuries CE,

for creating economic growth

at intermediate locations

between opposite ends of a trade route.

And I think the opportunity here

is to leverage the finite speed of light

to deploy infrastructure

in many locations around the world

that currently don't have access

to low latency Internet,

and as a result,

treat geography
as a new form of natural resource

that can lead to
broader economic development.

Thank you very much.

(Applause)