有我所不乐意的在你们将来的黄金世界里 我不愿去

THE PROBLEM

John Hebeler (0:02): The core problem is, our ability to create information has far exceeded our ability to manage it. It’s kind of like we’re drowning in our richness, that’s kind of what’s happening, cause you have all this data, all these access points, and there’s really no way to really help you deal with it except for stuff you can pull into your human brain. And you can only pull in so much. So you’ve got this massive amount of potential, but there’s not any real tools to harness it.

David Weinberger (0:33): We have so much stuff that we have to deal with. Individually, as a culture. So much – that it just bursts the bounds of any physical library. You know if we had a Dewey Decimal System for everything on the web, the trillion pages and all the subpages and all that, we wouldn’t find a thing, that system simply can’t work.

Footage of HE.Net Data Center: 800 cabinets = 9,600 terabytes = 9.6 billion thick books = 1,690 Libraries of Congress.

Clay Shirky (1:09): The amount of media that’s available to the average user is a vastly much larger superset than anything that’s ever existed in human history. If I was going to start a news business tomorrow, I would start a news business designed to produce not one new bit of news, but instead to aggregate news for individuals in ways that mattered to them.

Nova Spivack (1:32): Google really was more important as the web was in millions of pages. Now we’re entering a web that’s going to be billions – well, it already is – that’s going to be billions and billions of pages, and soon trillions of pages. Because a tweet is actually, every individual item is a page. Every product in the world, everything you can name or address is going to have a page. And so that’s trillions of things. And Google doesn’t scale to that.

Hebeler (1:59): There should be enough information out there that you should be able to ask for something extraordinarily specific, but you can’t. You pretty much have to do all the integration in your own head, you’ve gotta come back and see all the stuff that comes back from Google, and say, Oh, I wonder how I could ask that, cause this was kinda right but this was wrong…Oh, I see why it came back, came this out, that isn’t what I want though.”

Tim Berners-Lee (2:18): And so that’s not really a search, I think people use the word search to mean this sort of parachuting in, crossing your fingers, and hoping to land somewhere really good.

Chris Dixon (2:25): You know when you’re looking for a camera and you go to some place and there’s like ten thousand cameras and you’re overwhelmed, and sort of studies show that people are actually less likely to buy something when they’re overwhelmed by these things and less likely to actually be happy with what they buy afterward.

Weinberger (2:37): We have too many emails, so we start to tag them or label them, Gmail calls them labels. And we start to apply labels. And then we get, maybe we start to get hundreds of labels and we think, Oh jeez, now I gotta label my labels.

Hebeler (2:48): All the tweets and all the MySpace and you start to think, What if I could start to put things together in all that flow of information? And in order to do that, you need some structure.

Alon Halevy (2:59): It’s clear that something needs to be done with more structured data.

Dixon (3:03): Like all the information might be out there, it’s just if it’s indexed in a really inaccessible form, you know a lot of times it might as well not be out there, right?

Shirky (3:09): That is, in many ways, the problem of the age. Right, content, as it turns out, is not king.

Weinberger (3:15): We are always going to be filtering the filters that filter our filters. That filter our filters.

Hebeler (3:20): How do I find the right file? How do I know that all those files belong there?

Spivack (3:24): How do you integrate data?

Jason Shellen (3:25): How do I keep up with all these new sources of information?

Shirky (3:29): How do you filter things to create more value than you can currently get?

Hebeler (3:33): And that is what the Semantic Web could eventually promise to do.

THE VISION

In 1989, a physicist named Tim Berners-Lee invented something called the World Wide Web

Berners-Lee (3:49): I wanted to reframe the way we use information, the way we work together.

It made the Internet pretty popular…But Tim wasn’t finished.

Berners-Lee at TED, 2009:

Berners-Lee (4:03): Now, twenty years on, I want to ask your help in a new reframing. I want you to put your data on the web.

Berners-Lee (4:10): Okay, data is brown and boxy and boring and that’s how we think of it, isn’t it, “data”. But in fact, data is about our lives. You just, you log onto your social networking site, pick your favorite one, you say, this is my friend. Bing! Relationship. Data. You say, this photograph, Oh it’s about, it depicts this person, Bing! That’s data. Data, data, data…

Hebeler (4:30): The Semantic Web, at it’s lowest level, is just an expression of information, that’s all it is. So the, how the web works today, for the most part, is human to human. A human being puts something in some format, the computer is, all it knows about is formatting information. It knows it’s supposed to make this bold, it knows it’s supposed to underline this, the computer doesn’t know anything more than it’s just a bunch of bits. So semantics merely adds extra information to help you with the meaning of the information.

Spivack (4:56): It’s really just like transforming the web into something that’s a little bit more like a database…Trying to make it a lot easier to find stuff, because we have an understanding and an index of what’s out there.

Lee Feigenbaum (5:11): So you have specific data items, whether they’re books or songs or news articles or people. And linking them together. And with Semantic Web technologies, the links mean something.

Hebeler (5:31): It’s all about relationships, it’s about relationships of one string to another string, or one number to another number…And if I have enough of those relationships, I can start to build context, and context is what it’s all about…If I said any kind of word, it’s the context that surrounds the word that really gave you the meaning. What your brain has really done is connected that one word with all kinds of relationships. In a technical sense, all the Semantic Web does is start to give all these relationships.

Berners-Lee (6:12): If you look at the original proposal for the web, there are different shapes for different things, like people and documents. And there are arrows going between them and the arrows are labeled. Sort of this includes this, this describes this. So I think the idea of wanting to capture the meaning of the relationships, capture that actual data, has been there for ages.

Hebeler (6:42): We know that there’s a structure to this, there’s a structure to all the information on the Internet.

THE CRITICS

Shirky (7:00): The reason there aren’t too many criticisms of the Semantic Web yet is that it operates in its own bubble. I think I’m unusual in having, six or eight years ago, gone out with a set of opinions that said this isn’t working because it’s not a good idea and it’s never going to work.

Dixon (7:18): Semantic Web is a word that began with a technical meaning…Now that word has morphed into a marketing term that’s sort of abused and thrown around and so I would almost argue to the extent that it’s maybe not a useful word anymore.

Shellen (7:33): In terms of the Semantic Web, you know the idea is that everything is linked. I still like that idea. I think potentially that’s a Utopian idea to strive for.

Halevy (7:44): In an ideal world, yes. If everybody was trained in database and knowledge representation technology, that’s how we would do it.

Shirky (7:53): I’ve often joked that the Semantic Web is a witness protection program for AI researchers. That what the Semantic Web held out was the possibility that instead of making machines think like people, we could describe the world in terms that machines were good at thinking about. So we would switch from trying to build up brains in silicon, and instead rerender the actual world as information. And that gets very quickly to one of the deepest, you know, questions in all of Western philosophy, which is: Does the world make sense? Or do we make sense of the world? I don’t think you can unambiguously describe the world. I don’t think you can describe the world, or even large subdomains of the world, in a way that all observers or even most observers will agree with.

Heath (10:21): I think most of us in this room disagree with David, and I think we need to show him. Take a school analogy, and take him out to the playground and show him we can do much more.

Bernstein (10:32): So when you do parenting there are only two people fighting unless the grandparents are at home, right, and they’ll be fighting with you. But this is, you know, a whole community of what was it, five-hundred odd, six-hundred people, who are fighting about a baby called the Semantic Web.

Karger (10:45): Right, I mean, we could all just sort of sit back and do the work that we like to do and not care what everybody else is doing, but we’re believers in the potential of this Semantic Web thing, that some wonderful things can come out of it, and that makes us care how it’s pursued.

THE FUTURE

Berners-Lee (11:18): What’s the funny thing about the web is that it seeps in from the bottom. But for every person, they said, well, Tim, you know, what did you feel in 1993 when the web really exploded? And generally that meant, it was when I found out about it. Everybody, different people found out about the web at different times. Or different people had this ‘aha’ moment at different times.

Feigenbaum (11:44): I think the web, the World Wide Web, is a couple of different things. From a technology sense, it’s some extraordinary successful protocols and communication methods that mean that my web browser can go out to any web server in the world and get it back and show it to me. From a more social sense, the web is Facebook and MySpace, and blogs and news sites, right? And it’s all the things we do on the web. And I think it’s similar with the Semantic Web.

Spivack (11:20): The first step is evolution, the second step is revolution. When, you know, once there’s enough good content out there we can make some systems that can reason across the web and solve problems, answer complicated questions, make amazing discoveries and linkages between things. That’ll be cool. That is off in the future.

Weinberger (12:38): I have no idea what’s going to happen. But in terms of the openness of the web and our ability to access it and sort of the fundamental features of the Internet that made it the Internet, what happens to those features depends upon economics and politics and culture and technology. And it could easily change in radical ways through an invention that somebody in a garage is inventing now.

Berners-Lee (13:06): It’s a platform. Just like the web. The idea of it is not that it should promote one particular sort of application. Just as the Internet didn’t promote a particular application, so I could design the web on it without asking anyone’s permission. Same way, Semantic Web is sort of built on top of web, it should just allow you to build whatever you’d like on top of it. What we – at this conference, I think, people can’t imagine, because they’re trying to make it work so much, they’re not going to imagine what things people will be able to do with it once it’s working and it’s well-deployed.

Me (13:36): Do you think you can imagine?

Berners-Lee (13:39): Nope, I can’t. If we end up building all the things I can imagine we’ll have failed.