The inventor of the web, Tim Berners-Lee,
outlines his ideas for a more "intelligent" web in an interview with the BBC programme, Go Digital.
Presenter Tracey Logan quizzed Tim Berners-Lee
Go Digital: The worldwide web (WWW) transformed the internet from an academic reference tool to an everyday source of information as useful and almost as easy to use as the telephone.
Tim, take us back to the time when the web was little more than a twinkle in your eye. At the time, what dreams did you have for it and did you ever imagine that it would take off in the way it did?
Tim Berners-Lee: I suppose for the few people who had got that twinkle in their eye, it was a catchy twinkle. I think everybody had dreams for it and were involved in the development of it.
But there were so many things like that, so many grand projects you think of that don't turn out, or which don't end up taking off. There isn't enough money or people don't think they are going to work and so on, and the WWW could really have easily been like that.
Maybe people would have gone on using the older systems on the internet, things like gopher where you switch for existing information systems which weren't hypertext. But they worked and some people found them simpler to use
GD: You were really designing is as a tool for yourself, a bunch of scientists really, did you ever imagine that ordinary people might use it?
TBL: No, it was designed to be absolutely universal, in fact that was the thing really that made it work.
At Cern for example, in the lab where I worked, people had designed this documentation system and they said hey, we've got this great big universal documentation system and you've all got to put your documents into it and you've got to put them in this format, and you put them in this hierarchical system of classification.
Well that's telling people how to do their work and telling people how to think. So the thing about the web was it was really supposed to be universal and anything should go in there.
Point and click
GD: Those of who use the internet now take hyperlinks for granted, the fact that we can click on those special bits of text and go instantly to another document or web page containing more information. Back in the early 1990s, did people really understand what you were talking about?
TBL: They knew about clicking on links because the idea of hypertext and links had actually been invented some time ago. In fact it was 1945, I think. Vannevar Bush wrote a great paper about how it could be done. But he imagined it all being done using microfilms and electric sensors and mechanics because he didn't have computers and he didn't have the internet then, and then Ted Nelson invented the idea of hypertext.
In fact, at that point in 1989 there were lots of help programs where you could read which told you how to run an operating system and how to get to a magic word, and you'd be able to click on it.
But all these hypertext systems were all very centralised, they were all sort of complete works, beautifully made and finished and they had an index in the middle and everything went somewhere.
You could make a link to the name of another document somewhere and there was a possibility that if the other person didn't co-operate, if they took their website down then you would get in effect, what turned out to be the 404 which everybody knows now unfortunately.
The error 404 is, if you like, the bane of the web user's life, but it is also what makes the web work because all the systems before were very big and cumbersome and they had systems of making sure that every link had an end.
It was getting to the decentralised, and if you like, chaotic, anarchic social system for the web which I think people found difficult to grasp.
Drawing circles and arrows
GD: Now if the WWW was part one of the dream then you are quoted as saying that this latest development which you are in London to promote is part two. So what is it and what dreams do you have for it now?
TBL: The first hypertext programs I had, what I found was interesting was circles and arrows diagrams of our lives.
When you've got a white board or black board and you are trying to explain something to someone, or trying to design something, a new series, a new building, a new organisation, you start drawing circles and arrows between them.
These arrows can be about this person works for this person, it's about relationships between things. It can be family trees, it can be design trees, all kinds of things. I was interested in capturing this information and that's why I was interested in web-like things.
In fact when you start drawing circles and arrows, often people try to make it into a tree to help keep their minds straight, but really it's a web, things can connect in this world to all kinds of other things.
An earlier program I'd had before the web, way back in 1980, allowed you to say when you made a link, allowed you to say what sort of things were, so this is an organisation, this is a person and the relationship between them is an employee of that organisation.
So the nice things about that is if you have told the computer that, then the computer can do things like draw up organisational bar charts, it can find out all the dependencies, and it can start answering interesting questions for you.
GD: And this is the kind of thing that tends to be missing at the moment. When you do a web search, it doesn't tell you the meaning of the results the web search throws up. So this semantic web is actually finding meaning in web pages?
TBL: Yes, if you like, it turns out that lots of stuff in our lives actually have got very well defined meanings, like a calendar for example.
Supposing you go into a calendar program and you are going to book that you are going to be recording something at a particular time, then the calendar program knows the start time and the end time and it might know about the occasion and other people you've invited.
Now, that's good solid data that if the machine knows that, you should be able to ask it, who has been at more than one recording session with me, who is available on Tuesday, for example. Now that's just a question of going through all the bits of the calendar and picking them out and calendar programs do that sort of thing.
But it only has access to the data in the calendar program. Maybe you've got another thing like your address book which has got a lot more information and you've got another thing which is your bank statement that you download from the bank. That's got pretty well defined information, it's got what you spent money on and when.
Then you've got photographs you've taken. But suppose you take a photograph album on your computer and you want to see when you took those photographs, so you drag the photograph album onto the calendar, what happens?
Well in an ideal world the calendar program would say, what are these? I don't know what they are, but I tell you what they've got date and times on them so I can represent them on the calendar.
GD: So possibly using this semantic web, or this idea you could actually have a system that would be able to say, write for itself on the picture, taken on a beach in Brazil on the 14 October, and it was a sunny day?
TBL: You've got it, because the camera stores the time it was taken, but then you connect that up to what you were doing at the time and then it knows where you were.
Weather data is actually an excellent example of public data which really is something you can re-use again and again, if it was available to a machine.
GD: If you had an entire web that worked on this principle, you could have a digital organism that had a phenomenal amount of information and you have written that you're moving more in the direction then of an internet that can reason.
Now that may be slightly scary to some people, the idea of the internet as a giant brain. It's something maybe people have feared in the past - is that a stupid fear that people have or do you think it's reasonable?
TBL: Computers will become so powerful and there will be so many of them with so much storage that they will in fact be more powerful or as powerful as a brain and will be able to write a program which is a big brain.
And I think philosophically you can argue about it and spiritually you can argue about it, and I think in fact that may be true that you can make something as powerful as the brain, really whether you can make the algorithms to make it work like a brain is something else.
But that is a long way off and in fact that's not very meaningful for now at all. All I'm looking for now is just interoperability for data.