Tim Berners-Lee - MIT Ideaslab -    World Economic Forum 2010

This is not a Pecha-Kucha presentation. It is a prezi.
It has am emphasis on on relationships in a map rather than pictures.
It needs to be timed to the talk.
The plan is to make a video.


The ** indicate view points in the sequence on the prezi path. There are about 30 of them. Sometimes two view points follow each other in fairl rappid successsion, in order for the audience to understand the relationship.

The text in the script is not verbatim, but indicates the general gist of the presntation  at that point.

The plan is to make a 1280x768 video of the prezi, timed to the talk

TBL

__________________________________________

** intelligence

We have been looking at intelligence.

There are two varieties of intelligence we have looked at: Human intelligence and Machine intelligence.

** overall frame

I'm interested in what happens when we not only get people
and machines working  together, but we do it at scale.
Billions of people, billions of machines

Why? Here are two things I would like us to do.
We need to solve the huge common problems like HIV/AIDS, cancer, global warming.
I  hope we discover new forms of democracy which are fairer, more powerful, and more efficient, through the involvement of machines.


Let's look at where we have got to so far, and where we need to get.

We know
 
**  Human intelligence

there are things which only people can do, like
music and poetry.
We feel that we have a sense of understanding things 
which machines don't yet share.

** overall frame
That is human intelligence..

** Machine
on the other side
we have machine intelligence.

Machines can so things like drive a train or an elevator.
They can index text, recognize faces, and are staring to be able to translate language.
They can do a few things by themselves. 
But not much.

** overall Frame
We have for years of course been trying to use computers.
To do things where people and computers can work together.

** working together
When we prepare taxes, or schedule an event involving many calendars,
the computer is helping. 
But its helping in a way we wouldn't really describe as intelligence.

And it we are talking here about just one person and one computer:
not  large scale collaboration.

** overall frame

Now let's think about what happens at scale.

** Web scale
Web Scale.
There are around 100 billion web pages.
That's more than the number of neurons in your brain.

** web drawing

Here Marius Waltz imagines how it might be connected.

** web diagram

And here is a diagram actually made from some of the connections between some of the domains.  It doesn't show anything like the number of web pages.

What happens is each dot in the web is a person, or a computer, trying
to contribute to making the world a better place?

** overall frame

Let's look first at what's happened recently as many people 
have started to connect their human intelligence across the web.

** Social network frame

Some of the most popular ways in which people interact nowadays
are based on social networking web sites.

Is this going to be the future of common intelligence?

There are two things wrong with this picture.

** facebook cartoon closeup

One is the fact that, as David Simonds expressed in this cartoon from the Economist,
the social networking sites tend to be silos.
You tell one site who your friends are, and then you go to a photo site.
You want to see photos of your friends, but you have to tell each new site all over again.
The information you give each site is hoarded.
It doesn't connect.
This isn't very web-like.

** Social network frame

The other is that the people using facebook 
are only using in a small way the intelligence
of the computer.
The computer may look at their behavior and figure
out who to suggest as friends, and figure out what
advertising they should see.
But the computer is not helping them solve
major problems.

So one of the things we need to do is to take these social networking situations
and make the connections in them part of the web of data
which machines can work with.

** overall frame

So social networking is one  the phenomena which have happened as people have connected using the web.
What happens when we connect machines?


** Connected machines

Well, not a lot, if we just connect them.
But we can program them to handle very large amounts of computer.
To search for signals from extraterrestrial intelligence.
To predict the weather.
In these cases they are not working an a weblike way: there are no emergent properties.

To expose, together, large amounts of data to the world as Linked Open Data.
In this case we get a lot of benefit from connecting data in different databases.

The systems though, in which  machines produce a lot of processing power
deal with data in on lump, 
not concerned about where each bit comes from
or what it is appropriate ti use it for.

So a direction we are moving at MIT 
is to build systems which are aware of social policies,
so they can bring machine power to a world where
provenance of and trust in information, are a function of the social networks.

** overall frame

So .. what do we need, to get humans and machines working together?
To tie all this intelligence into something which is much more powerful?

** connections

Intelligence is about connections.


** Intelligence is all about connections

The storage and manipulation of connections.

** shared between humans and machines

Knowledge shared between people and machines. 

** connections


** half-formed idea

One of the most challenging connections 
is that involve in global creativity.
How to we connect the half-formed idea in my mine to the half-formed idea
in yours, to make a solution -- even if you are across the world?

** connections

Connections are stored on the web.
It is stored on the web as links.
This is semantic web technology.


** represented ion the web by 

** Links

This is the semantic web technology.
It what is already being used by an exploding set of public open databases, linked together and forming a nucleus for a growing web of shared knowledge.

* * overall    @@

But it is is in the very early stages. 
The more things move into this space, the more value there is for
everything else in this space.
And the closer we are to the next paradigm shift in which 
people and machines really work together at scale.
So let's review how we can move things into that space.
These are some of the directions we are taking at MIT.

**webize

Where we have computers helping people, but not at scale, we need
to webize them: make these applications part of the a global network of knowledge.

** add semantics
As we have seen, there are areas in which we have got two of the three ingredients.

Where we have people woking globally using the web, we need to expose the network of connections they are making to machines.  We need to add semantics.

** Policy-aware

Where we have connected machines providing power, we need to add an awareness of the social aspect,
and make them policy-aware so that they can really work responsibly in our world.

** overall

It is a powerful vision, and it is coming to fruition in small areas. But meanwhile we
still do not understand the this huge system -- the web -- humanity and machines connected -- behaves.  So we are also creating a new field of Web Science. 
But that is another story.

Thank you