My name is Tim Berners-Lee, I invented the world wide web when I was working in Switzerland and Geneva. Now we are here at MIT, in the computer science and artificial intelligence lab, CSail, in a new building, the Stata Center which has just been opened.

Here I do a number of things. I am the director of the World Wide Web consortium, which is a consortium of companies which looks at all the ways here which the web could be better, more stable, and could evolve to be more powerful or exciting. Also, I’m a senior research scientist at MIT. We have a research group looking further ahead at the same things, and making sure that the things which the consortium will take on in a few years actually make sense.

INVENTING THE WORLD WIDE WEB

When I say I invented the web, I really just put together the last few pieces out of a set of construction kit, which had already been made. So I was looking at a problem of there being lots and lots of different sorts of systems for keeping information and there was already hypertext systems and there was already the internet. The internet had spread across America, and it had just gotten into Europe, so when I was looking at the problem of all these different information systems being incompatible, so you couldn’t easily get information from one to the other, I realized that we could make them all look like one big virtual information system just by taking the ideas of hypertext on one side and then using internet protocols to connect all the computers, we would make what I call, for better or worse, World Wide Web: W W W.

The computers in fact were actually linked together already. People were using email already and but what’s different about the web is that whereas before you could get information from a computer by logging on to it, being very aware of what was happening and running lots of different programs on different machines - the idea of the web is that the computers get out of the way, and they leave person looking at the screen, looking at information, but being in an information space where really the computer is no longer visible, the computer is the hidden servant, just getting all the information across the network but in a way so that all the user thinks about is a click, following a link from one page to another. And so the person browsing the web can be creative, can be thinking about information as opposed to be thinking the darn computer.

Two things were very important about the way the web was designed: One was that it’s completely universal, you can put anything in it; the other is that it’s decentralized in its architecture. Decentralized means that there’s no central computer that everybody has to connect to; it also means that there’s no central organization that they have to connect to, be involved with, or subscribe to, or pay money to. The idea of this is that the web should be very much of a grass roots phenomenon, that is should be able to grow by people building little webs and then connecting those webs at the edges, not by having been originally designed as part of a huge master plan. And that’s important because society forms all sorts of different organizations, people connect in all sorts of different ways in order to solve problems, and often when you’re looking in a particular place very logically for a problem, really it turns out that a random connection, a random association in your mind or in the information you find, will take you to the solution. So it’s important for the web to be able to handle those random associations. Anything must be able to link to anything. So that was important property about the web, and that was built into the web architecture, and in a way does connect with the way the mind works. The human mind is amazingly good at remembering a smell of a place, or the background sound of a place and then bringing to mind the place when you smell the same coffee in a totally different place. And it’s that sort of ability which gives the mind the ability to jump across and suddenly pluck the solution out from somewhere completely different to where you were looking.

ARTIFICIAL INTELLIGENCE

I don’t think computers will be smart in the sense that human beings are smart and have the same sort of intelligence for a long time because I think we’ve been, as a race, trying to make machines smart for a long time, and its been always more difficult than we thought. However, theoretically, there’s no reason why you shouldn’t make a very very big very fast computer which can have more and more of the abilities to deal with things in the way people do, or to mimic the way people think. In fact, of course, the great thing about computers is that they can do things that we can’t do, that they are good at handling matters of numbers, so in a way, certainly what’s exciting now is using computers for things that they are good at and leaving people to do the things that people are good at.

THE IMPACT OF THE WEB

Since the web has come - and in developed countries has become fairly ubiquitous - what’s new is not that we can do more work, or we can read more stuff; what’s new is that we have a greater choice. So what’s changed is that a person can go to their computer and go to anywhere in the world, can get involved in almost any activity that they are interested in. So it means that you can do anything. But of course you can’t do everything. So it means that a person has got a much more difficult choice. The great thing is I think that because different people will get involved in all sorts of different things, hopefully they we will start breaking out of our boxes. That if we can do ten things in one life then those ten things will not be all in the same area, we will be involved in a few different countries perhaps. We are involved in different fields, and by doing that we’ll make bridges, we’ll take ideas, we’ll help people think in the same way, we’ll help people understand each other, and hopefully we’ll not only solve big scientific problems but also, by this understanding, we’ll help people live together in different countries in peace.

Well, first of all its not clear that I would have been an extremely wealthy man if I’d have patented a thing or two, because there were lots of small hypertext projects which didn’t take off the ground and many of them because there was some degree of central control. If I’d of been the central person that you’d have to register every click with, pay for every click, or register every web page then I would have had a business model. But business wouldn’t have existed; there would be no web out there. Instead lots of people would have made lots of incompatible webs, they would have tried to avoid my patent, so they would have made systems which didn’t work the same way. They were kind of similar, attack the problem from different angles. And the whole value of a web page, the fact that that link can in principle go to anything, that is the wonderful thing that makes the web the web, and it wouldn’t have happened, and we wouldn’t be talking about it now.

HOW THE WEB WORKS

To understand how the web works, you have to understand that the internet is something different from the web. The Internet is a system which existed well before I invented the web, which connects all computers. The way the internet works is that when one computer wants to send a message to another computer, it takes the message as a series of numbers with each character written out with a number in a special code, so code equaled unicode. And when it takes that series of numbers it breaks them into little boxes called packets, and on each block first it puts, like the address on a postcard, it outs the number of the computer that that block’s got to go to. So that it takes this packet, as it’s called, and sends it down the wire that goes to the next computer. Now these computers are arranged in a network, they really are like a fishing net, so they’re connected so that the packet can go lots of different ways, but the next computer looks at the number of computers it’s going to and decides which way to send it best on its route to its destination. So each packet makes its way across the network, and eventually, hopefully arrives at the far end. If it gets lost on the way, as sometimes happens, then, well, a copy will be sent again. In this way, one computer can send to another computer a message. So that all existed before the web.

Now the way the web works is that when you’re looking at a screen, you’re looking at a document, a web page, you can see a certain amount of text and pictures, but when there’s a link there’s something the computer can see that you can’t see. Behind the link, there is a URL, for example http://www.something/somethingorother. And that long thing, which people nowadays write on buses and you see on adds, originally that was designed to be just a hidden part of the hypertext system, that’s why its such a complicated set of characters. What happens when you click on the link is that your computer looks up the URL, looks at the first piece of it and say oh http. I know how to get a document using http. So then it looks at the next piece which is www.something, that is the name of the service somewhere on the internet which is going to provide it with a document it’s looking for. First thing it does, it sends a message off to the domain name service, which will say, what’s the computer number? Because remember that packets go across with their address on them is a number. So it sends a message to a special computer, which is the nearest local domain name server, sends back a message saying the number you asked for is this, this is the computer number. So now your computer has got the computer number of the computer which has got the next web page that you need, so the computer makes a little message saying, “get,” and it puts the URL of the page its looking for: http://www.thinking/something/something/somethingelse and it says, I want to get that, puts that in a packet and sends that all the way across the internet and so that is routed from computer to computer until it gets to the web server, to the computer which is responsible for serving up the information from that particular web site. When that computer gets the message, it looks up what that piece of information is and sends it back to your computer, and when it comes back, it’s a message with a whole lot of characters with all the web page, and its not just the character that you see, also the little angle brackets. In between angle brackets there’s the coded up stuff that the computer uses, the coded up stuff to say which of this should be the heading, which of this should be the image, which of this should be small texts, which texts of this is going to be a footnote and so on. But also, which text of this should be a link, and in between those angle brackets the coded stuff that again you don’t see as the computer shows you the new web page, on that new web page there are all the links and again the computer has made a mental note of the URL’s for all the links you could take from that web page. So you read some of the web page, you figure out where you want to go and you click, and it starts all over again.

Hypertext is a word invented by Ted Nelson who used it to mean non-linear text. He was fascinated by the idea of breaking out of text as it is in an article or a paper where you just read on and on and on. He liked the idea of being able to jump out to something which happened to be more relevant to you if you were reading with a particular interest in mind. So that word, hypertext, we use for text with links. Text where instead of just reading on the page you can follow a link and go somewhere else.

The World Wide Web is a project to implement one big global hypertext system. The way the World Wide Web works is that every page has got a name, and that name for historical reasons is called a URL or URI, that means Universal Resource Identifier or Locator. It’s the name of the document. It’s universal because you can give one to anything; you can make URI’s for all kinds of things. R is for Resource, meaning just something of value out there, so something out there on the web; it could be movie, it could be some data, it could be the text; it could be something about the weather. These are what we call resources. And Locator is used if you think of it as taking you to the place in the web where something is. Identifier we use if you think of it as being more of a name for something. In fact, the way the web works, it’s both. It’s the name and it’s also a name that allows you to actually seem to get there as you navigate through the web.

When I invented it, it was called a UDI, a Universal Document Identifier. And Document was changed to Resource because it was felt that it was too constraining a term.

Well we still have a lot of philosophical discussions about exactly what a resource is. And the architecture of the web is something which we are trying to explain to people so that as it’s added onto so rapidly, people don’t break it, people don’t use the system in a way that prevents other parts of it from working properly. And the question as to exactly how that web operation works, when for example one computer is getting some logical information about a piece of data from another computer instead of a person browsing the web, is still important to nail down precisely.

THE WORLD WIDE WEB CONSORTIUM

I invented the web back in, I suppose, 1989 I wrote the first memo. And then during the next few years there were just a few volunteers all over the planet who got together over the internet, exchanging email, exchanging news group messages, and exchanging software to make the thing a reality. And it took off in fact very slowly because it was difficult to get everybody to realize that it was going to be a good idea, difficult to explain what it was going to be like. But by 1993-1994 it was really taking off and the companies involved wanted to get together to give it some stability to be able to plan, to have somewhere to meet to talk about the future, so we formed the World Wide Web consortium; that’s w3.org. And if you go to the consortium’s web site www.w3.org then you’ll see on that a list of a huge number of ways in which the web is now evolving. People are browsing the web not only by using screens now all sorts of shapes and sizes, they’re using their wrist watches, they’re talking, they’re using a phone to talk to a program which is going to guide them through ordering a pizza, or guiding them through ordering a flight and a hotel, or booking a car, or by if you like taking a voyage not through the visible set of hyperlinks but through a set of possible conversations which are now making more of a web. So voice XML is a technology which allows that; people are using web technologies within an enterprise to allow computers to talk to each other and we’re looking at the whole area of putting information, which is more like data than documents, it’s more the sort of thing you’d want to process with a spreadsheet or a database program than print out and read. But that’s important information too, the web must be universal so we’re looking at data on the web, and we call that the semantic web.

We’ve talked about the web from the point of view of somebody browsing the web, but originally when I built it, what I really wanted was some place where people could work together. In other words where not only could you read stuff but you could write at the same time. So the original web program in fact, it was like a browser except that you could open a new window and write a new web page any time you wanted and you could very easily with a few key strokes you could highlight a piece of the thing that you were writing and it would make a link to something that you had just been reading. So you could build a web of links very easily and then you could just hit another hot key and everything would be saved back and you’d have a new set of connected thoughts. What I found was really exciting about that was the idea of people sharing a hypertext web. People working together to build a common understanding, for example how they were going to build something. I was working as a software engineer, so my problem was getting people to work together to make a piece of software, and I found they were always having to pass on information from people who just left the project to people who were just joining the project and the documentation would tend to get lost and was in little pieces. The idea of being able to link it all together, to really explain to anybody where we are now. And to make the process of design a collaborative thing where you actually capture that collaborative process was really important. So the original web browser was also an editor. However as the other web browsers were built by other people, none of them had this capability to write and the web became, for one reason or another, a medium in which people edited things offline and then uploaded them. So this business of downloading and editing and uploading became a rather clumsy process, easier than producing a newspaper perhaps, and a lot of people were very excited about the fact that to make something like a newspaper or an Op Ed column like a ‘blog’ was much faster nowadays. But it still hasn’t really for most people become a collaborative space, so one of the ways in which we’re trying to make the web go bit by bit is to become something where an average person can make more than bookmarks. At the moment people send email and they bookmark things. Some of the recent developments in ‘wikies’ and ‘blogs’ are really interesting. I think the fact that these wikies, which are web pages where you can click on a button and get a form for editing the page, even though they’re very very crude, they’re very popular. And I think they’re popular because people really need to have spaces where they can collaboratively edit things.

Wikie Wikie, quick, Wikie I think is Hawaian for quick. Wikie is a type of web site where you can very quickly edit it; its short and cheap and cheerful. It’s not very sophisticated, you can’t make complicated mark-up. It does not for example allow you to control who has access in most cases, so wikies rely on a lot of trust. If you write something, somebody else can go change it. You have to do it a spirit of collaboration where you only override what somebody has written if you know that they’d be okay with it.. We spent a lot of time trying to connect the two.

Originally the web was designed to be a collaborative space; clearly as a universal space its going to have some things like company brochures and the tax law, and the front pages of major national dailies which go through a lot of editing processes, which is really a one-way process where there are a huge number of people involved, getting the quality right, it goes through a bottleneck of the editing process. And then when it’s been published, it then is available to a lot of people, and that’s one model, but it’s only one model. Another model is the scribbling note space that I keep, the sort of spiral bound notebook equivalent. We should have that on the web as well, and we should have a continuum so I can write scribbly notes in my spiral bound notebook and those can mature as I share them with more and more people through the equivalent of a blog to the equivalent of a published paper, maybe a newspaper article or maybe a scholarly publication. The web must be universal space and span all these different types information.

In the World Wide Web consortium, we looked at what was happening and we saw progress on many fronts, but we found that this collaborative space wasn’t really happening. So we wondered what we could do, the first thing we did was try to use it for ourselves. So we built a web server where you could save things back, and one of the first things that showed us was that really people weren’t prepared to collaborate unless they really could control which groups they were in and who had access to it. So yes, who can read my paper? Who can write my paper? These are things which are really important to me. I’m not going to play in the space unless I feel I’ve got a pretty good idea over those things, so we found we had to develop the access control and the social systems for building groups, allowing people to easily build groups, determine who was in them, and then connect the documents to the groups, which documents had what sort of access in which sort of groups. And the collaborative space really is made of a web of different groups, talking to each other, knowing that whenever they talk to each other it will take more time, but on the other hand they will have wider agreement. Making social decisions about how big the group should be and how much effort they are going to spend working with which other groups.

THE FUTURE OF THE WEB

Exactly what form the applications will take in the future isn’t clear at all. We’ve had some fairly clear application boundaries; we’ve had writing programs and collaborative programs, and email, and instant messaging. But I think that the boundaries between these are rather artificial, and challenging those boundaries is always important. I think it’s important that the software environment should be something where people can invent new forms of program which may combine these things, fit together in other ways and introduce them. When I wrote the World Wide Web program people could write other programs like it, which communicated across the internet and just run them. It’s very important that we should continue in that spirit of being able to make your applications. We should continue to play with what really works, what helps us communicate more easily with each other and also what helps us communicate more easily with the computers which can help us with the difficult more numerical bureaucratic tasks.

I think our future must be a more collaborative sort of writing. Something which spans all the boundaries. And then after they’ve all said it seems reasonable then we might make it public. Currently, if you look at the blog software, it only deals really with writing something and then making it public. If you look at Wikies software, it requires a lot of trust from a particular group. I think there is going to be many years of exploration, experimentation, and programmers desperately trying to use their skills to make something that works better.

BOOKS

Physical books I love, and I think the physical book is by no means dead. As I said there is a spectrum of things and there is a certain area there when something has had a lot of effort put into it and I’ts been published. There’s nothing like printing it out, and putting it in your back pocket and going down to the beach, or keeping it for quick reference. I think there is something very nice about the physical medium of a book. I also think there is something very special about places of learning and places of storage of information. When you walk into a library it’s actually not just the fact that you may find the book you want there. When you walk into a library, you are walking into somewhere which has been put there as a storage of knowledge and has been put there as a place of learning. And I think anybody who goes into an institute of education, or a place which has been built as a library is immediately swept up into that. Same sort of thing with a physical book that this physical embodiment of the message of the author I think is very powerful, and we are in a physical place; physical places are very important, they will always be important, as will physical books.

Well, I think that most of the use of computers will just be the tools, so people used to say, what are you doing now? Oh, I’m searching the web. When it was new people would think of surfing the web as an occupation. But it’s not. Now it’s, what are you doing now? Oh, I’m doing my homework. What are you doing now? Oh, I’m looking to buy a new refrigerator. And the tool is a laptop which is sitting on the table along with a pen and paper, along with a cup of coffee, along with various other tools of doing, so that the web of information becomes part of what’s life; we know how to use it, it’s integrated, and I’ts just one of the many tools we use. The communication through the medium of a computer keyboard and screen is part of a communication with other people, which we use as well as meeting them face to face and talking to them on the phone and so on.

I think that there is a huge amount of what you can call “the record,” what Vannavar Bush called the record of the knowledge of human kind, that which is public, that which is history, that which is being the product of a lot of effort over the ages which is great and very important to make available to as many people as possible. I think that we’ve got to work very hard before people in developing countries will have the same access as people in developed countries, and I think that we also have to make sure that we capture things and make them available, easily make sure there aren’t commercial, legal restrictions put on them which somehow mean that you’re not getting the whole story. So that is true for a huge body of work, which will be a very large effort to put on line. That doesn’t mean that everybody should have everything direct access. That doesn’t mean that everybody should have access to everything in the sense of: I should have access to your calendar; I should know what you wrote in your journal last night; I don’t have to have access to the inter-workings of a committee which hasn’t published its recommendations yet because the committee will work better eventually if they can talk behind closed doors before they come out into the open. Families are very important units, we respect them by giving them their privacy. People are really important, and it’s just us people, so we must respect individual people’s privacy, their rights to think, to write privately, and to read privately without everybody knowing what they’ve chosen to read today. So part of the building of this society is building the barriers as well in those areas. It is the barriers which actually protect people, protect groups, and make them actually exist.

E-MAIL

Sometimes I hope that the web has undone some of the damage that television has done in making us less literate, with all due respect to the medium. I think one of the nice things about email is that people communicate using the written word, and the written word is expressive in a different way to television. Television can show you things that you can’t see on a black and white printed sheet, but a black and white printed sheet can make you imagine things that you could never see with television. They are both very powerful. One of the nice things I think about email is that a lot of people in fact turn out to be very literary, and to write really powerful and beautiful emails, and to express themselves extremely eloquently just using the written word. So I think the internet, the written word, is alive and well and is flourishing on the internet as well as off the internet.

Emails with bad spelling in fact are a form, they are a genre, of communication. Just as when you pick up a sheet of paper, if it’s glossy, it’s got the company logo on the top; then you imagine the company is going to be standing by that piece of paper. But if you pick up something which has been rushed off or has memo confidential on it, you know that probably it might change or you may not get agreement from everybody. There are different forms of communication. When people send email with bad spelling, in it can be a message to say: this is a rough, this is a hand scribbled note equivalent. So there are lots of these questions of using short-handed email, there are questions of using these little abbreviations which not only save time, but also they show, just take this as what I’m sending you off the top of my head. I want to let you know that I haven’t thought this through; don’t quote me on this, this is just my initial reaction. If I send you something which I want you to quote to the press, you’ll know because I’ll say so, and it will be spell checked.

COPYRIGHT

One of the social systems which we haven’t finished thinking through is how we reward people for their creativity. When the books were, if you like, the license, when I bought the rights to read something by buying a physical copy of it, when it was really difficult to make a copy of that, when even if I had made a carbon copy it would be really much less pleasant to read simply because of the physical limitations; now we’ve got no physical limitation, we’ve broken through that. What we can do is think up new ways of connecting the reader and the writer so that the reader can compensate the writer, and I think there are lots of ways you could do that. And just taking the old laws and pushing them violently in one way or another is not really appropriate I think. We could be much more creative; we also have to be careful. We don’t want to mess up the systems which already exist, we don’t want to stifle creativity. So when we publish a book on the net, the danger is the systems aren’t set up in the same way. Now if you publish music on the net, you can publish it in a way that people can easily buy copies, pay good money for it in such a way that they can share it with their friends but not with the entire planet. So there are more and more systems being put together so that people can do the right thing, so that people can be honest without being forced into using a particular way of buying things. There’s a choice in how to publish things and there’s a choice as to how you buy music. I think that the choice of systems is really important because I think it will be a long time before we’ve finished exploring all the new social systems we could set up for making this work well.

Currently, for example, you can download audible books as music, you can do that for music, you can download your own copy for your own enjoyment. Currently you can download books on tape by paying for them, you can download them onto a little mp3 player which you carry around in your pocket and the money will go back to the publishing company, and a certain amount of it will go back to the author.

It’s a big area as I said... there are lots of different modes and styles and what I think, to take one particular one and focus on it is inappropriate because they will be continually changing.

A UNIVERSAL AND OPEN SPACE

I think the important things that we’ve said, we’ve talked about the importance of collaborative, we’ve talked about how the web works basically, we’ve talked about the importance of social systems around the technical systems and that how we should engineer those very carefully. If we are going to end on a note, I suppose about the web, the important thing is its universality. The web is built as a universal medium. Whatever we do in the future, it’s very important we don’t alienate any particular type of information. We shouldn’t alienate data because it’s only understandable by programs, we shouldn’t alienate information from a particular culture of information that uses a particular character set, or that’s read right to left instead of left to right. We shouldn’t alienate information which is scribbled and start setting some sort of standards for what constitutes quality information. We should also do our very best to make sure that it’s accessible by people in different countries, and also that it’s accessible by people who may have disabilities, who may use different senses; where you or I might read the written word on a screen, somebody else might have to listen to it, or they need to read it on a special screen with it blown up using a special font and a lot of magnification. This universality, the fact that information of all different types is being reused in all different ways, including being reused in ways that its original creators didn’t understand is very, very important to the web. So a lot of the work at the consortium and the work of everybody who is developing the web should be to respect that universality and to try to do whatever they can to make sure it remains a universal and open space.