I want to talk to you about two of the most exciting possible things. You've probably guessed what they are -- data and history. Right? So, I'm not a historian. I'm not going to give you a definition of history. But let's think instead of history within a framework.
So, when we're making history, or when we're creating historical documents, we're taking things that have happened in the past, and we're stitching them together into a story. So let me start with a little bit of my own story. Like anybody my age who works creatively with computers, I was a popular, socially well-adjusted young man --
(Laughter)
And sporty! Sporty young man. And like a lot of people my age in the type of business that I'm in, I was influenced tremendously by Apple. But notice my choice of logo here, right? The Apple on the left, not the Apple on the right. I'm influenced as much by the Apple on the right as the next person, but the Apple on the left -- I mean, look at that logo! It's a rainbow. It's not even in the right order!
(Laughter)
That's how crazy Apple was.
(Laughter)
But I don't want to talk too much about the company. I'll start talking about a machine, though. How amazing it is to think about this. I go back and I think about this. Wednesday -- one Wednesday, when I was about 12 years old, I didn't have a computer. On Thursday, I had a computer. Can you imagine that change? It's so drastic. I can't even think about anything that could change our lives that way. But I'm actually not even going to talk about the computer. I'm going to talk about a program that came loaded on that computer. And it was build by, not the guy on the left, but the guy on the right. Does anybody know who the guy on the right is? Nobody ever knows the answer to this question. This is Bill Atkinson. And Bill Atkinson was responsible for tons of things that you see on your computer every day. But I want to talk about one program that Bill Atkinson wrote, called HyperCard. Someone's cheering over there.
(Laughter)
HyperCard was a program that shipped with the Mac, and it was designed for users of the computer to make programs on their computers. Crazy idea today. And these programs were not the apps that we think about today, with their large budgets and their big distribution. These were small things, people making applications to keep track of their local basketball team scores or to organize their research or to teach people about classical music or to calculate weird astronomical dates. And then, of course, there were some art projects. This is my favorite one. It's called "If Monks Had Macs," and it's a nonlinear kind of exploratory environment. I thank the stars for HyperCard all of the time. And I thank the stars for putting me in this era where I got to use HyperCard. HyperCard was the last program to ship on a public computer that was designed for the users of the computer to make programs with it. If you talked to the people who invented the computer and you told them there would be a day, a magical day, when everybody had a computer but none of them knew how to program, they would think you were crazy.
So let's skip forward a few years. I'm starting my career as an artist, and I'm building things with my computer, small-scale things, investigating things like the growth systems of plants. Or, in this example, I'm building a simulated economy in which pixels are trading color with one another, trying to investigate how these types of systems work, and just kind of having fun. And then this project led me to start working with data. So I'm building graphics like this, which compare "communism" -- the frequency of usage of the word "communism" in the New York Times -- to "terrorism," at the top. You see "terrorism" kind of appears as "communism" is going away. And with these graphics, I was really interested in the aesthetic of the graphs. This is Iran and Iraq. It reads like a clock. It's called a "timepiece graph." This is another timepiece graph, overlaying "despair" over "hope." And there's only three times -- actually, it's "crisis" over "hope" -- there's only three times when "crisis" eclipses "hope." We're in the middle of one of them right now. But don't think about that too much.
(Laughter)
And finally, the culmination of this work with the New York Times data a few years ago was the attempt to combine an entire year's news cycle into a single graphic. So these graphics actually show us a full year of news, all the people, and how they're connected into a single graphic.
And from there, I started to be interested again in more active systems. Here's a project called "Just Landed," where I'm looking at people tweeting on Twitter. "Hey! I just landed in Hawaii!" -- you know, how people just casually try to sneak that into their Twitter conversation. "I'm not showing off. Really. But I did just land in Hawaii." And then I'm plotting those people's trips, in the hopes that maybe we can use social network and the data that it leaves behind to provide a model of how people move, which would be valuable to epidemiologists, among other people. And, more fun -- this is a similar project, looking at people saying "Good morning" to each other all around the world. Which taught me, by the way, that it is true that people in Vancouver on the West Coast wake up much later and say "Good morning" much later than the people on the East Coast, who are more adventurous. Here's a more useful -- maybe -- project, where I took all the information from the Kepler Project and tried to put it into some visual form that made sense to me.
And I should say that everything I've shown you up to now -- these are all things that I just did for fun. It may seem weird, but this comes back from HyperCard. I'm building tools for myself. I may share them with a few other people, but they're for fun, they're for me. So, all these tools I show you kind of occupy this weird space somewhere between science, art and design. That's where my practice lies. And still today, from my experience with HyperCard, what I'm doing is building visual tools to help me understand systems.
So today, I work at the New York Times. I'm the data artist in residence at the New York Times. And I've had an opportunity at the Times to work on a variety of really interesting projects, two of which I'm going to share with you today. The first one, I've been working on in conjunction with Mark Hansen. Mark Hansen is a professor of statistics at UCLA. He's also a media artist. And Mark came to the Times with a very interesting question to what may seem like an obvious problem: When people share content on the internet, how does that content get from person A to person B? Or maybe, person A to person B to person C to person D? We know that people share content in the internet, but what we don't know is what happens in that gap between one person to the other. So we decided to build the tool to explore that, and this tool is called Cascade.
If we look at these systems that start with one event that leads to other events, we call that structure a cascade. And these cascades actually happen over time. So we can model these things over time. Now, the New York Times has a lot of people who share our content, so the cascades do not look like that one, they look more like this. Here's a typical cascade. At the bottom left, the very first event. And then as people are sharing the content from one person to another, we go up in the Y axis, degrees of separation, and over on the X axis, for time. So we're able to look at that conversation in a couple of different views: this one, which shows us the threads of conversation, and this one, which combines that stacked view with a view that lets us see the threads.
Now, the Times publishes about 7,000 pieces of content every month. So it was important for us, when we were building this tool, to make it an exploratory one, so that people could dig through this vast terrain of data. I think of it as a vehicle that we're giving people to traverse this really big terrain of data. So here's what it really looks like, and here's the cascade playing in real time. I have to say, this was a tremendous moment. We had been working with canned data, fake data, for so long, that when we saw this for the first moment, it was like an archaeologist who had just dusted off these dinosaur bones. We discovered this thing, and we were seeing it for the first time, these sharing structures that underlie the internet. And maybe the dinosaur analogy is a good one, because we're actually making some probabilistic guesses about how these things link. We're looking at some of these pieces and making some guesses, but we try to make sure that those are as statistically rigorous as possible.
Now tweets, in this case, they become parts of stories. They become parts of narratives. So we are building histories here, but they're very short-term histories. And sometimes these very large cascades are the most interesting ones, but sometimes the small ones are also interesting. This is one of my favorites. We call this the "Rabbi Cascade." It's a conversation amongst rabbis about this article in the New York Times, about the fact that religious workers don't get a lot of time off. I guess Saturdays and Sundays are bad days for them to take off. So, in this cascade, there's a group of rabbis having a conversation about a New York Times story. One of them has the best Twitter name ever -- he's called "The Velveteen Rabbi."
(Laughter)
But we would have never found this if it weren't for this exploratory tool. This would just be sitting somewhere, and we would have never been able to see that. But this exercise of taking single pieces of information and building narrative structures, building histories out of them, I find tremendously interesting.
You know, I moved to New York about two years ago. And in New York, everybody has a story that surrounds this tremendously impactful event that happened on September 11 of 2001. And my own story with September 11 has really become a more intricate one, because I spent a great deal of time working on a piece of the 9/11 Memorial in Manhattan. The central idea about the 9/11 Memorial is that the names in the memorial are not laid out in alphabetical order or chronological order, but instead, they're laid out in a way in which the relationships between the people who were killed are embodied in the memorial. Brothers are placed next to brothers, coworkers are placed together. So this memorial actually considers all of these myriad connections that were part of these people's lives.
I worked with a company called Local Projects to work on an algorithm and a software tool to help the architects build the layout for the memorial: almost 3,000 names and almost 1,500 of these adjacency requests, these requests for connection -- so a very dense story, a very dense narrative, that becomes an embodied part of this memorial. Working with Jake Barton, we produce the software tool, which allows the architects to, first of all, generate a layout that satisfied all of those adjacency requests, but then second, make little adjustments where they needed to to tell the stories that they wanted to tell. So this memorial, I think, has an incredibly timely concept in our era of social networks, because these networks -- these real-life networks that make up people's lives -- are actually embodied inside of the memorial. And one of the most tremendously moving experiences is to go to the memorial and see how these people are placed next to each other, so that this memorial is representing their own lives.
How does this affect our lives? Well, I don't know if you remember, but in the spring, there was a controversy, because it was discovered that on the iPhone and, actually, on your computer, we were storing a tremendous amount of the location data. So Apple responded, saying, this was not location data about you, it was location data about wireless networks that were in the area where you are. So it's not about you, but it's about where you are.
(Laughter)
This is very valuable data. It's like gold to researchers, this human-mobility data. So we thought, "Man! How many people have iPhones?" How many of you have iPhones? So in this room, we have this tremendous database of location data that researchers would really, really like.
So we built this system called Open Paths, which lets people upload their iPhone data and broker relationships with researchers to share that data, to donate that data to people that can actually put it to use. Open Paths was a great success as a prototype. We received thousands of data sets, and we built this interface which allows people to actually see their lives unfolding from these traces that are left behind on your devices. Now, what we didn't expect was how moving this experience would be. When I uploaded my data, I thought, "Big deal. I know where I live. I know where I work. What am I going to see here?" Well, it turns out, what I saw was that moment I got off the plane to start my new life in New York; the restaurant where I had Thai food that first night, thinking about this new experience of being in New York; the day that I met my girlfriend. This is LaGuardia airport.
(Laughter)
This is this Thai restaurant on Amsterdam Avenue. This is the moment I met my girlfriend.
See how that changes the first time I told you about those stories and the second time I told you about those stories? Because what we do in the tool, inadvertently, is we put these pieces of data into a human context. And by placing data into a human context, it gains meaning. And I think this is tremendously, tremendously important, because these are our histories that are being stored on these devices. And by thinking about them that way, putting them in a human context -- first of all, what we do with our own data is get a better understanding of the type of information that we're sharing. But if we can do this with other data, if we can put data into a human context, I think we can change a lot of things, because it builds, automatically, empathy for the people involved in these systems. And that, in turn, results in a fundamental respect, which, I believe, is missing in a large part of technology, when we start to deal with issues like privacy, by understanding that these numbers are not just numbers, but instead they're attached, tethered to, pieces of the real world. They carry weight. By understanding that, the dialog becomes a lot different. How many of you have ever clicked a button that enables a third party to access your location data on your phone? Lots of you. So the third party is the developer, the second party is Apple. The only party that never gets access to this information is the first party! And I think that's because we think about these pieces of data in this stranded, abstract way. We don't put them into a context which, I think, makes them a lot more important.
So what I'm asking you to do is really simple: start to think about data in a human context. It doesn't really take anything. When you read stock prices, think about them in a human context. When you think about mortgage reports, think about them in a human context. There's no doubt that big data is big business. There's an industry being developed here. Think about how well we've done in previous industries that we've developed involving resources. Not very well at all. I think part of that problem is, we've had a lack of participation in these dialogues from multiple pieces of human society.
So the other thing that I'm asking for is an inclusion in this dialogue from artists, from poets, from writers -- from people who can bring a human element into this discussion. Because I believe that this world of data is going to be transformative for us. And unlike our attempts with the resource industry and our attempts with the financial industry, by bringing the human element into this story, I think we can take it to tremendous places.
Thank you.
(Applause)
