sponsored links
TED2013

Eric Berlow and Sean Gourley: Mapping ideas worth spreading

February 28, 2013

What do 24,000 ideas look like? Ecologist Eric Berlow and physicist Sean Gourley apply algorithms to the entire archive of TEDx Talks, taking us on a stimulating visual tour to show how ideas connect globally.

Sean Gourley - Physicist and military theorist
Sean Gourley, trained as a physicist, has turned his scientific mind to analyzing data about a messier topic: modern war and conflict. He is a TED Fellow. Full bio

sponsored links
Double-click the English subtitles below to play the video.
Eric Berlow: I'm an ecologist, and Sean's a physicist,
00:12
and we both study complex networks.
00:15
And we met a couple years ago when we discovered
00:17
that we had both given a short TED Talk
00:19
about the ecology of war,
00:21
and we realized that we were connected
00:23
by the ideas we shared before we ever met.
00:25
And then we thought, you know, there are thousands
00:27
of other talks out there, especially TEDx Talks,
00:29
that are popping up all over the world.
00:31
How are they connected,
00:33
and what does that global conversation look like?
00:34
So Sean's going to tell you a little bit about how we did that.
00:36
Sean Gourley: Exactly. So we took 24,000 TEDx Talks
00:39
from around the world, 147 different countries,
00:43
and we took these talks and we wanted to find
00:46
the mathematical structures that underly
00:48
the ideas behind them.
00:50
And we wanted to do that so we could see how
00:52
they connected with each other.
00:53
And so, of course, if you're going to do this kind of stuff,
00:55
you need a lot of data.
00:57
So the data that you've got is a great thing called YouTube,
00:58
and we can go down and basically pull
01:02
all the open information from YouTube,
01:03
all the comments, all the views, who's watching it,
01:06
where are they watching it, what are they saying in the comments.
01:08
But we can also pull up, using speech-to-text translation,
01:11
we can pull the entire transcript,
01:14
and that works even for people with kind of funny accents like myself.
01:16
So we can take their transcript
01:19
and actually do some pretty cool things.
01:21
We can take natural language processing algorithms
01:23
to kind of read through with a computer, line by line,
01:25
extracting key concepts from this.
01:28
And we take those key concepts and they sort of form
01:30
this mathematical structure of an idea.
01:33
And we call that the meme-ome.
01:36
And the meme-ome, you know, quite simply,
01:38
is the mathematics that underlies an idea,
01:40
and we can do some pretty interesting analysis with it,
01:43
which I want to share with you now.
01:44
So each idea has its own meme-ome,
01:46
and each idea is unique with that,
01:49
but of course, ideas, they borrow from each other,
01:51
they kind of steal sometimes,
01:53
and they certainly build on each other,
01:54
and we can go through mathematically
01:56
and take the meme-ome from one talk
01:58
and compare it to the meme-ome from every other talk,
02:00
and if there's a similarity between the two of them,
02:02
we can create a link and represent that as a graph,
02:04
just like Eric and I are connected.
02:07
So that's theory, that's great.
02:10
Let's see how it works in actual practice.
02:11
So what we've got here now is the global footprint
02:14
of all the TEDx Talks over the last four years
02:16
exploding out around the world
02:19
from New York all the way down to little old New Zealand in the corner.
02:20
And what we did on this is we analyzed the top 25 percent of these,
02:24
and we started to see where the connections occurred,
02:27
where they connected with each other.
02:30
Cameron Russell talking about image and beauty
02:31
connected over into Europe.
02:33
We've got a bigger conversation about Israel and Palestine
02:35
radiating outwards from the Middle East.
02:37
And we've got something a little broader
02:40
like big data with a truly global footprint
02:41
reminiscent of a conversation
02:43
that is happening everywhere.
02:45
So from this, we kind of run up against the limits
02:47
of what we can actually do with a geographic projection,
02:49
but luckily, computer technology allows us to go out
02:52
into multidimensional space.
02:54
So we can take in our network projection
02:55
and apply a physics engine to this,
02:57
and the similar talks kind of smash together,
02:59
and the different ones fly apart,
03:01
and what we're left with is something quite beautiful.
03:03
EB: So I want to just point out here that every node is a talk,
03:05
they're linked if they share similar ideas,
03:08
and that comes from a machine reading
03:11
of entire talk transcripts,
03:13
and then all these topics that pop out,
03:15
they're not from tags and keywords.
03:17
They come from the network structure
03:19
of interconnected ideas. Keep going.
03:21
SG: Absolutely. So I got a little quick on that,
03:23
but he's going to slow me down.
03:25
We've got education connected to storytelling
03:26
triangulated next to social media.
03:28
You've got, of course, the human brain right next to healthcare,
03:30
which you might expect,
03:32
but also you've got video games, which is sort of adjacent,
03:34
as those two spaces interface with each other.
03:36
But I want to take you into one cluster
03:39
that's particularly important to me, and that's the environment.
03:40
And I want to kind of zoom in on that
03:43
and see if we can get a little more resolution.
03:45
So as we go in here, what we start to see,
03:47
apply the physics engine again,
03:49
we see what's one conversation
03:51
is actually composed of many smaller ones.
03:53
The structure starts to emerge
03:55
where we see a kind of fractal behavior
03:57
of the words and the language that we use
03:59
to describe the things that are important to us
04:01
all around this world.
04:03
So you've got food economy and local food at the top,
04:04
you've got greenhouse gases, solar and nuclear waste.
04:06
What you're getting is a range of smaller conversations,
04:09
each connected to each other through the ideas
04:12
and the language they share,
04:14
creating a broader concept of the environment.
04:15
And of course, from here, we can go
04:18
and zoom in and see, well, what are young people looking at?
04:19
And they're looking at energy technology and nuclear fusion.
04:23
This is their kind of resonance
04:25
for the conversation around the environment.
04:27
If we split along gender lines,
04:29
we can see females resonating heavily
04:31
with food economy, but also out there in hope and optimism.
04:33
And so there's a lot of exciting stuff we can do here,
04:37
and I'll throw to Eric for the next part.
04:39
EB: Yeah, I mean, just to point out here,
04:41
you cannot get this kind of perspective
04:43
from a simple tag search on YouTube.
04:44
Let's now zoom back out to the entire global conversation
04:47
out of environment, and look at all the talks together.
04:52
Now often, when we're faced with this amount of content,
04:54
we do a couple of things to simplify it.
04:57
We might just say, well,
05:00
what are the most popular talks out there?
05:01
And a few rise to the surface.
05:04
There's a talk about gratitude.
05:05
There's another one about personal health and nutrition.
05:07
And of course, there's got to be one about porn, right?
05:10
And so then we might say, well, gratitude, that was last year.
05:13
What's trending now? What's the popular talk now?
05:16
And we can see that the new, emerging, top trending topic
05:19
is about digital privacy.
05:22
So this is great. It simplifies things.
05:25
But there's so much creative content
05:27
that's just buried at the bottom.
05:28
And I hate that. How do we bubble stuff up to the surface
05:30
that's maybe really creative and interesting?
05:34
Well, we can go back to the network structure of ideas
05:36
to do that.
05:39
Remember, it's that network structure
05:40
that is creating these emergent topics,
05:43
and let's say we could take two of them,
05:45
like cities and genetics, and say, well, are there any talks
05:46
that creatively bridge these two really different disciplines.
05:49
And that's -- Essentially, this kind of creative remix
05:52
is one of the hallmarks of innovation.
05:54
Well here's one by Jessica Green
05:56
about the microbial ecology of buildings.
05:58
It's literally defining a new field.
06:00
And we could go back to those topics and say, well,
06:02
what talks are central to those conversations?
06:04
In the cities cluster, one of the most central
06:07
was one by Mitch Joachim about ecological cities,
06:09
and in the genetics cluster,
06:13
we have a talk about synthetic biology by Craig Venter.
06:14
These are talks that are linking many talks within their discipline.
06:18
We could go the other direction and say, well,
06:21
what are talks that are broadly synthesizing
06:23
a lot of different kinds of fields.
06:25
We used a measure of ecological diversity to get this.
06:26
Like, a talk by Steven Pinker on the history of violence,
06:29
very synthetic.
06:32
And then, of course, there are talks that are so unique
06:33
they're kind of out in the stratosphere, in their own special place,
06:35
and we call that the Colleen Flanagan index.
06:38
And if you don't know Colleen, she's an artist,
06:41
and I asked her, "Well, what's it like out there
06:44
in the stratosphere of our idea space?"
06:45
And apparently it smells like bacon.
06:47
I wouldn't know.
06:50
So we're using these network motifs
06:52
to find talks that are unique,
06:54
ones that are creatively synthesizing a lot of different fields,
06:55
ones that are central to their topic,
06:58
and ones that are really creatively bridging disparate fields.
07:00
Okay? We never would have found those with our obsession
07:03
with what's trending now.
07:05
And all of this comes from the architecture of complexity,
07:07
or the patterns of how things are connected.
07:10
SG: So that's exactly right.
07:13
We've got ourselves in a world
07:15
that's massively complex,
07:17
and we've been using algorithms to kind of filter it down
07:19
so we can navigate through it.
07:22
And those algorithms, whilst being kind of useful,
07:24
are also very, very narrow, and we can do better than that,
07:26
because we can realize that their complexity is not random.
07:30
It has mathematical structure,
07:33
and we can use that mathematical structure
07:34
to go and explore things like the world of ideas
07:36
to see what's being said, to see what's not being said,
07:38
and to be a little bit more human
07:41
and, hopefully, a little smarter.
07:43
Thank you.
07:45
(Applause)
07:46

sponsored links

Eric Berlow - Ecologist
TED Senior Fellow Eric Berlow studies ecology and networks, exposing the interconnectedness of our ecosystems with climate change, government, corporations and more.

Why you should listen

Eric Berlow is an ecologist and network scientist who specializes in not specializing. A TED Senior Fellow, Berlow is recognized for his research on food webs and ecological networks and for creative approaches to complex problems. He was the founding director of the University of California's first environmental science center inside Yosemite National Park, where he continues to develop data-driven approaches to managing natural ecosystems. 

In 2012 Berlow founded Vibrant Data Labs, which builds tools to use data for social good. Berlow's current projects range from helping spark an egalitarian personal data economy to protecting endangered amphibians in Yosemite to crowd-sourcing novel insights about human creativity. Berlow holds a Ph.D. from Oregon State University in marine ecology.

 

 

Sean Gourley - Physicist and military theorist
Sean Gourley, trained as a physicist, has turned his scientific mind to analyzing data about a messier topic: modern war and conflict. He is a TED Fellow.

Why you should listen

Sean Gourley's twin passions are physics (working on nanoscale blue-light lasers and self-assembled quantum nanowires) and politics (he once ran for a national elected office back home in New Zealand).

A Rhodes scholar, he's spent the past five years working at Oxford on complex adaptive systems and collective intelligent systems -- basically, using data to understand the nature of human conflict. As he puts it, "This research has taken me all over the world from the Pentagon, to the House of Lords, the United Nations and most recently to Iraq". Originally from New Zealand, he now lives in San Francisco, where he is the co-founder and CTO of Quid which is building a global intelligence platform. He's a 2009 TED Fellow.

In December 2009, Gourley and his team's research was published in the scientific journal Nature. He is co-founder and CTO of Quid.

sponsored links

If you need translations, you can install "Google Translate" extension into your Chrome Browser.
Furthermore, you can change playback rate by installing "Video Speed Controller" extension.

Data provided by TED.

This website is owned and operated by Tokyo English Network.
The developer's blog is here.