[This is an edited/corrected ver­sion of the tran­script post­ed by the World Wide Web Foundation (licensed CC-BY). Unfortunately, only part of the orig­i­nal video record­ing is avail­able, so the cut­off is indi­cat­ed below and only min­i­mal edits were made from there on.]

Well, thank you, Alberto, for those words. Thank you for invit­ing us to the din­ner. Thank you for your sup­port. I sup­pose the path to my stand­ing here traces its way at one point back in March to a con­ver­sa­tion I had with Gary Kebbel at anoth­er din­ner with a lot few­er peo­ple. And at that din­ner, we were just chat­ting across a small table about our own hopes and fears. And it turned out that Gary’s hopes and fears for jour­nal­ism, and I sup­pose maybe the hopes and fears of the Knight Foundation through his eyes, had an awful lot of over­lap with the things that I felt about the Web. And of course a lot of jour­nal­ism hap­pens on the Web, and an impor­tant part of the Web is journalism.

And we were wor­ried. We were con­cerned about, with so much infor­ma­tion how do you tell where some­thing came from? How do you tell what’s good and bad? What’s reli­able infor­ma­tion or not? What is commercially-biased infor­ma­tion? We had con­cerns about how, if you are going to label infor­ma­tion good and bad, who’s going to start mak­ing the deci­sions in the new world? How can you build mer­i­toc­ra­cies which, some­how, select those people?

In gen­er­al, I sup­pose, the big ques­tion for journalism—the big ques­tion for the Web—does it sup­port democ­ra­cy? What does it take to sup­port democ­ra­cy? So it turns out that we were both get­ting very excit­ed about these two things, and indi­rect­ly that led to my com­ing here. But that’s not what I came to talk to you about. I came to talk to you about think­ing about the future of the Web, very much more broad­ly, and see if I can enlist your help in that.

But I’ll start by going back to the begin­ning of the Web. In fact let’s go a lit­tle bit fur­ther back, 1989. I’m a pro­gram­mer. I’m work­ing at CERN. I’m a soft­ware engi­neer. I’m build­ing pieces of soft­ware to work in that huge accel­er­a­tor, that huge col­lid­er, that they had built. CERN is on the French-Swiss bor­der; the Collider is under­ground. At that time, the big col­lid­er was the Large Electron-Positron Collider. Now you may have heard in the news recent­ly about the Large Hadron Collider that they’ve just turned on. Well, the Large Hadron Collider, they just built in the tun­nel which they made for the pre­vi­ous one. And as it hap­pens, back in 1989, in fact they were just start­ing that project. Okay. That’s 1989.

I wrote a memo point­ing out that with the huge amount of infor­ma­tion around at CERN that you had to keep track off, all those big pieces of— You may have seen on tele­vi­sion because of the Large Hadron Collider, that there are huge pieces of equip­ment there. And they had [been] built by teams from all over the plan­et, and these teams come togeth­er using dif­fer­ent com­put­er sys­tems, using dif­fer­ent soft­ware, dif­fer­ent doc­u­men­ta­tion for­mats. They of course speak dif­fer­ent lan­guages, but they try at CERN to restrict them­selves to a mix­ture of English and French. But if you’re sit­ting there as a soft­ware engi­neer try­ing to make head and tail of it, you feel you real­ly need an inte­grat­ed infor­ma­tion sys­tem. Because you have to go and con­nect to this com­put­er for that piece of infor­ma­tion, and this com­put­er for that piece of infor­ma­tion. And when you con­nect to it you have to learn a new pro­gram on each computer.

So that was the sit­u­a­tion I was in. And I wrote a memo say­ing, ​“We should have a glob­al hyper­text sys­tem to fix this.” The memo, I dis­trib­uted it to a few peo­ple but there’s nowhere real­ly to dis­trib­ute it to at CERN because CERN is a physics lab. It did­n’t have com­mit­tees for build­ing pro­grams and hyper­text systems.

So what hap­pened was basi­cal­ly noth­ing for eigh­teen months. But because the old lab in fact was more or less com­plete and the new LHC had­n’t real­ly ramped up, they weren’t real­ly quite get­ting into the swing of cre­at­ing that, there was a bit of a lull. And dur­ing that lull, my boss Mike Sendal said at one point, ​“Well, why don’t you— Tim, why don’t you try out this new com­put­er you’ve got, this new fan­cy NeXT big black box you’ve got­ten? Why don’t you try it out with some­thing, devel­op some­thing? Why don’t you – you need some pro­gram just to see how good it is as a plat­form for devel­op­ing things on? Why don’t you do that do that hyper­text thing you were talk­ing about?” So, okay! ​“Sure, Mike.”

So I went away and start­ed mid­dle of October and start­ed pro­gram­ming. So I wrote a thing which you would prob­a­bly rec­og­nize as a Web brows­er. Actually it was a browser/editor. It would allow you to cre­ate Web pages, too, because I thought that actu­al­ly being able to cre­ate was just as impor­tant as being able to read.

When I cre­at­ed the program—there’s a nice lit­tle sort of pro­gram for gen­er­at­ing program[s] on that machine. You can say I want to write a new pro­gram and the first thing it asks you is, ​“Okay, what’s the name of it?” Typed in ​“World Wide Web”. I’d thought about it a bit. I’d reject­ed a few oth­er things, but you know ​“World Wide Web” kind of came to the top of the pile of ​“infor­ma­tion mine” and things, and ​“mine of infor­ma­tion.” They were…well, it became…for bet­ter or worse I decid­ed to call it the ​“World Wide Web.” 

I wrote the pro­gram, wrote a Web serv­er, and defined more impor­tant­ly the lan­guages which those spoke. The com­put­er lan­guage for send­ing Web pages across the Internet, which I called HTML, and the pro­to­col which hap­pens between the two com­put­ers, the Web serv­er and the Web client, called HTTP, and so on. And these URLs, which you know these things start ​“http” and all that.

And that was all tech­ni­cal. That was all the tech­ni­cal bits of invent­ing the Web. But of course, to the design of the Web there’s a lot that isn’t tech­ni­cal. There’s a lot that’s social. Because when some­body makes a link, when a link is cre­at­ed from your blog, you link to some­body else’s blog, well, it’s a per­son that makes the link. It’s a per­son that fol­lows the link if they want to. So, real­ly, the fact that the whole Web thing works is because peo­ple are moti­vat­ed. And it’s real­ly a sys­tem of peo­ple. Yes, it’s enabled by technology.

So as time went on, we stopped think­ing about the Internet as a col­lec­tion of com­put­ers, and we stopped think­ing about the Web as a col­lec­tion of con­nect­ed pages. And now we think of the Web as human­i­ty con­nect­ed. There’s always been an impor­tant social aspect to it. In fact, there were a lot of social con­straints behind its growth. For exam­ple, ear­ly on there was a cru­cial point— At that point… Does any­body remem­ber Gopher? Who remem­bers the Internet Gopher? One, two, three—yes! Okay, I know you do. I know you. But not very many peo­ple remem­ber the Internet Gopher.

Well folks, the rea­son that the rest of you don’t remem­ber the Internet Gopher is the Internet Gopher, it was a won­der­ful sys­tem. It was very like the Web and in fact it was grow­ing expo­nen­tial­ly like this when the Web was grow­ing expo­nen­tial­ly like this. [ges­tures to indi­cate two ris­ing slopes, the sec­ond at a slow­er rate] But at a cer­tain point the University of Minnesota, which had pro­duced the Gopher sys­tem, sent out a lit­tle email that went around which said that there was just a remote pos­si­bil­i­ty that pos­si­bly, if you were a com­mer­cial per­son, maybe if you were run­ning a serv­er, they might charge a very small fee for using the system.

So at con­fer­ences, at one par­tic­u­lar con­fer­ence I remem­ber, I was just besieged by peo­ple in the cor­ri­dor say­ing, ​“We have dropped. We are not doing any­thing. You know, I’ve been doing all these for free in my garage, why should I work for the University of Minnesota in my spare time?” Or, ​“I work for a big com­pa­ny. My lawyers won’t allow me to touch the code because now this is a pro­pri­etary sys­tem. What’s the sto­ry with the World Wide Web?”

So actu­al­ly, I did­n’t have a sto­ry then. My col­league, Robert Cailliau and I, it took us eigh­teen months to per­suade CERN that the tech­nol­o­gy behind the World Wide Web should be royalty-free for any­body to use. But it hap­pened, and that was real­ly impo— So, a few months lat­er I could go back to them and say, ​“You’re safe. You can work on it.” And then Gopher, you could look at the charts of the amount of Internet traf­fic. Gopher… [hand ges­ture indi­cat­ing a sud­den drop]. Web… [hand ges­ture indi­cat­ing a quick­ly ris­ing slope] And the rest, well, it was expo­nen­tial. The expo­nen­tials con­tin­ued in a fright­en­ing fash­ion. But if CERN had not made that real­ly impor­tant social step of say­ing that the Web would be royalty-free, we would not be here talk­ing about it. The Internet would be a very dif­fer­ent place.

So that was an impor­tant step in the evo­lu­tion of the World Wide Web. And anoth­er impor­tant step hap­pened a few years lat­er. The first few years, I had to per­suade a lot of peo­ple that the Web would­n’t be too com­pli­cat­ed, would be a good idea. I had to try to find peo­ple to take the idea that I’d had and I’d imple­ment­ed on this NeXT machine— Anybody remem­ber the black NeXT machine? Yeah? Anybody still got one? Yeah. Yeah, okay. So I need­ed it [to] run on PC and I need­ed it to run on Mac. And peo­ple across the world actu­al­ly vol­un­teered and start­ed to write browsers. And the famous one of course was the one that Marc Andreessen wrote, NCSA. 

But as more and more peo­ple start­ed writ­ing Web browsers, they start­ed mak­ing them more fan­cy. They added fea­tures. At first, the orig­i­nal HTML did­n’t have pic­tures, did­n’t have tables. So they added tables. But when you put a table into your Web page, you had to do it dif­fer­ent­ly depend­ing on whether you want­ed to read it in this brows­er or that brows­er. And now, this was clear­ly going to be the end of the Web as we know it. The real­ly impor­tant point about the Web is it’s one Web. When you make a hyper­text link from a Web page, you can poten­tial­ly make it to absolute­ly any­thing on the Web. Everything on the Web has got a URI. (One of those things start­ing ​“http”; URL if you like.) You can take it and you can use that, make a link, and you could make that thing link to any­where. So it’s one big sys­tem. That is cru­cial. If it had bro­ken into pieces, it would­n’t have worked. And there was a threat that it was going to break into pieces. Everybody was wor­ried. Everybody want­ed to have tables, but they want­ed tables in the same form.

Something unusu­al hap­pened at one point. I was this pro­gram­mer at CERN. CERN is a big cam­pus. Lots of peo­ple wan­der­ing around look­ing like…you know, stu­dent physi­cists, pro­gram­mers, kind of very un-corporate. But I got a call from the CERN recep­tion, which it does have one. And these cor­po­rate types turned up. They came from the American com­put­er com­pa­ny, Digital Equipment Corporation. They had suits, and they want­ed to see me.

So I found some­thing which looked like a con­fer­ence room and put them in it, and found what they want­ed. And they were the team that was plan­ning Digital Equipment Corporation’s Internet strat­e­gy. And they were plan­ning to turn the com­pa­ny around and make that com­pa­ny real­ly com­plete­ly reset itself, giv­en the Internet.

And so they want­ed to know, if we’re going to be bas­ing every­thing on this Web thing, and it’s based on all these spec­i­fi­ca­tions which you’ve got on a disk some­where in your office (sort of won­der­ing which disk it was), then how can we make sure it’s a lit­tle more sta­ble? How can we make it evolve? How can we make sure that our engi­neers can get their new ideas out into it? How can we make sure that as it expands, as it evolves, it does every­thing that we want, but it stays one Web? And one of them sug­gest­ed in fact that I should put togeth­er a con­sor­tium, maybe that MIT might agree to host it as a neu­tral body. 

Well in fact, to cut a long sto­ry short yeah, they did. MIT were hap­py to host it as they’d host­ed con­sor­tia before. And also they were hap­py for it not to just be an American thing and for us to have a base in Japan and a base in Europe. 

So that was 1994, was when the World Wide Web Consortium start­ed. And since then it’s been a neu­tral body, and it’s been a very impor­tant part of the Web because it’s where all these com­pet­ing ven­dors of Web tech­nol­o­gy come to agree on the stan­dards. And with­out the stan­dards we just would­n’t have one Web. The world gets by with hav­ing a dif­fer­ent plug on elec­tri­cal things. You kin­da get by, any­way. It’s a pain. But you know, you can get by by hav­ing a dif­fer­ent plug and all these kind of things that you take when you go abroad to plug into the pow­er some­where else. It can’t get by if all the Web servers from a par­tic­u­lar com­pa­ny speak one lan­guage, and all the Web browsers from anoth­er com­pa­ny speak a dif­fer­ent one. It has to be one Web. The Consortium was a real­ly impor­tant part of putting that togeth­er. So that was anoth­er impor­tant step in the evo­lu­tion, set­ting up the Consortium to set the standards.

[Available record­ing ends here]

And as time went on, Consortium ran with a steady tech­ni­cal staff, about six­ty peo­ple, keep­ing this going and var­i­ous new tech­nolo­gies came in and became deployed. Some work bet­ter than oth­ers, some are great tech­nolo­gies but not com­plete­ly deployed on all browsers like Scalable Vector Graphics, great tech­nol­o­gy deployed on all browsers but one, almost there.

But as time went on, it was clear that we had to do more than just roll out the stan­dards. Because stan­dards typ­i­cal­ly are agree­ing about doing things we know how to do. You don’t make a stan­dard nor­mal­ly until it’s some­thing you know how to do. What about the things we don’t know how to do? Shouldn’t we be hav­ing stu­dents think more broad­ly? Shouldn’t we be think­ing about the Web a bit fur­ther out?

And in fact, think­ing about the Web, the Web itself now a few years on, it’s not a col­lec­tion of Web servers. Yes, there was a time I had a list of Web servers. I’ve got a copy that when I had twenty-six entries on it. But it was a lot more than a very large list of Web servers. It was some­thing so big nobody could keep the whole list.

And even the com­pa­nies that did try to crawl over the whole Web and keep huge index­es of it, they found that actu­al­ly has a very com­pli­cat­ed struc­ture. It has some real­ly, real­ly strong nodes which have huge num­bers of links point­ing to them. It has the famous Long Tail of lots and lots and lots of very small sites. It has lit­tle com­mu­ni­ties. Communities that some­times are geo­graph­i­cal, some­times not, has a com­pli­cat­ed structure.

Every now and again it has a new phe­nom­e­non starts. The wiki changes the way we cre­ate com­mu­nal infor­ma­tion in a way that I would­n’t have guessed would have worked. And I think a lot of peo­ple who have guessed would­n’t work. But we have Wikipedia which in gen­er­al is not com­plete non­sense. And we have blogs, which you may think are com­plete non­sense except for your blog of course.

But that again pro­duced a new genre. So we have these new things crop­ping up and we don’t – but the Web itself con­nect­ed human­i­ty with all its com­plex­i­ty, its size, there are more – about the same num­ber rough­ly of web­pages out there as there are neu­rons in your brain. The bad news is the num­ber of web­pages is increas­ing. [Laughter]

So, it’s big and it’s com­pli­cat­ed and for your brain we have cog­ni­tive sci­ence which brings togeth­er lots of dif­fer­ent dis­ci­plines to look at the brain for we decid­ed we need web sites. So we start­ed talk­ing about web sites three years ago. And we cre­at­ed a Web Science Research Initiative in fact, ini­ti­at­ed by some peo­ple from Southampton University in the UK and MIT but lots of oth­er peo­ple join­ing in. And web sites now, three years on, we’ve had work­shops to try and define new chal­lenges. We’ve got stu­dent exchange pro­grams. We want to get peo­ple to col­lab­o­rate togeth­er and it’s start­ing to hap­pen. That is in fact com­ing out in Scientific American, there’s a nice arti­cle on Web Science you might want to look at.

So, Web Science now adds to the stan­dards. We now have a sep­a­rate but impor­tant­ly con­nect­ed body of effort going on towards the sci­ence, not the sci­ence of study­ing it but of course also build­ing it, for engi­neer­ing it. Science in the sense of com­put­er sci­ence, writ­ing pro­grams too.

So we have the stan­dards and we have the sci­ence. And it was actu­al­ly, to sort of the segue into the next piece, for me was part­ly when we were sit­ting down at the Web Science Research Initiative, defin­ing as one does the goal for the orga­ni­za­tion. And we decid­ed that prob­a­bly the best way to define the goal for Web Science was that the Web should some­how opti­mal­ly serve human­i­ty. That the Web should ser­vice human­i­ty. That we should study the Web, devel­op it, and check that it is serv­ing humanity.

Well, you know, when you’re work­ing on stan­dards and you’re work­ing on sci­ence, you’re doing research and you’ve got that goal that the Web should serve human­i­ty. You can’t just sit there and do research about what’s going to put the next lat­est tech­nol­o­gy into the hands of the exec­u­tive, in their PDA on the exec­u­tive chair. Or the high-powered sci­en­tist who’s going to use the Web on their desk to dis­cov­er course to disease.

When you talk about human­i­ty, you can’t eth­i­cal­ly ignore the fact that 80%, I have heard, of the world don’t use the Web. There are peo­ple who can’t use the Web because they just don’t have any con­nec­tiv­i­ty. But I think also about the peo­ple who actu­al­ly like if 80% of the world who do have sig­nal, they could if they had the right device, they had a plan and the device had some sort of a Web brows­er on them that worked. And there was some­thing use­ful on the Web which would work for them and the infor­ma­tion was actu­al­ly in their native tongue. And they were lit­er­ate or it pro­vid­ed some­way of get­ting infor­ma­tion which worked with some­body who was illit­er­ate. And if the source of things they want to do, the sort of social net­work­ing they want­ed to do actu­al­ly worked with their cul­ture in the sort of net­work­ing they want to do, then those peo­ple would be able to use it.

So as we devel­op the Web, we should­n’t just aim to make it the most fan­cy thing but we should also have a third part which is being very aware of the needs of soci­ety as a whole. So there is a third piece.

Now, I must admit, I had some con­cerns about this. Because I felt that it is pos­si­ble that when you’re look­ing at a devel­op­ing coun­try, real­ly what they need is health­care. They need food. They need clean water and who are we as tech­nol­o­gists to come along, who are we who should sug­gest that they need our Internet, the thing that we play with. And I express this in fact, sit­ting down on a log by a lake, talk­ing to a mis­sion­ary. Somebody who’d been to out­er Africa and who had done lot of work, work­ing with peo­ple who had been in war-torn areas, recov­er­ing from the ter­ri­ble effects of war.

And so, I sug­gest­ed that maybe we should leave it and wait until there was water in there. And she said, ​“Well,” all she had to do was tell me one sto­ry of one per­son. This per­son, this man had taught him­self English by using var­i­ous books that he’d come across. Among them are the bible I under­stand, some­thing that he could get in both lan­guages. Having taught him­self English and hav­ing got some sort of rudi­men­ta­ry Internet con­nec­tion, then he could sell his ser­vices as a trans­la­tor on the open mar­ket out there in the big wide world. He could not only – so he could bring mon­ey into the vil­lage but also he could be a chan­nel to bring infor­ma­tion into the vil­lage. He could trans­late the stuff that was out there into the vil­lage. And he could oper­ate – so he could enable commerce.

So, there are a lot of anec­dotes we’ve heard about both ways. I think we real­ly have to lis­ten to the sto­ries, go out there, under­stand and talk to every­body who’s got expe­ri­ence about it. To make sure that we real­ly do under­stand what the needs are before we make any rash assump­tions about the fact that – assum­ing that they would want the same sort of thing which we would want.

But, the results of this then of step­ping back is, if you like, for me the next phase in the evo­lu­tion of the Web. The next piece of, if you like, of duty. Of the thing that we feel we have to do that we have an oblig­a­tion to do. And that is that there is the Web for all of soci­ety. There is the sci­ence and the standards.

And so, it is my plea­sure and hon­or to unveil this evening at this din­ner that we will be cre­at­ing a Web Foundation. The Web Foundation will coor­di­nate these three things. It will be unique in that it will be able to use the expe­ri­ence to influ­ence the research. Use the research to sug­gest to peo­ple in the under­served areas what might be pos­si­ble. To let them think out­side the box and demand some­thing bet­ter. And use both of these to feed into the process that makes the stan­dards, that make the new pro­to­cols for the Web.

So, the Web Foundation’s mis­sion is to advance one Web. One Web which is free and open. I don’t need to talk to you about the First Amendment. I talked in a lot of oth­er places about Net neu­tral­i­ty, the open­ness of the Web. The fact that I can choose what source and infor­ma­tion I take is very important.

So, it will advance one Web that’s free and open. And it will extend the capa­bil­i­ties of the Web. The capa­bil­i­ties and also the robust­ness of the Web. Can we trust this thing? It will still be there in a few years time. How do we know? And also, it will extend the ben­e­fit of the Web to every­body on the plan­et. It will try to. It will do it’s best to do these three things. [Applause]

So, I men­tioned I’m grate­ful to being here. While I’m grate­ful for the fact that the Knight Foundation in fact have giv­en us a $200,000 plan­ning grant in order to start this work to start fig­ur­ing out how to put togeth­er the Foundation. Steve Bratt, my col­league at the World Wide Consortium has guar­an­teed, has vol­un­teered to take on the job of lead­ing the Web Foundation. Certainly in these ear­ly stages as it takes off.

You know this is not a launch. This is just, as I say the start­ing of the plan­ning. We expect a launch as we put it togeth­er with our ini­tial found­ing donors. We expect a launch at the begin­ning of next year. In the mean­time, watch this space. Well actu­al­ly, watch www​.web​foun​da​tion​.org, a domain name to which we actu­al­ly got hold of a cou­ple of days ago. Before that, we had to have a dash in it. So now, you can use all one word.

So, the Web Foundation we’re putting togeth­er, it’s in a way it feels like a very big under­tak­ing. Because the things that we hope for, for the Web, well in a way, we want the Web to sup­port human­i­ty. Of course, we have huge hopes for human­i­ty. Many of them in this build­ing the hopes of infor­ma­tion ser­vice to human­i­ty you see are sort of etched into var­i­ous pieces of rock around this build­ing. The hope that peo­ple will be able to build on the Web.

But it’s all this ques­tion of build­ing on the Web you noticed? It’s not that we’re going to design the Web to do a par­tic­u­lar thing for a par­tic­u­lar appli­ca­tion. The point about the Web is it’s a plat­form. It should be, for the next gen­er­a­tion, for the peo­ple who are stu­dents now, for peo­ple who are chil­dren now, they should find that the Web is a can­vas that they can draw won­der­ful things on. If some­body asks me, ​“OK, so what’s going to be the result in ten years? What do you imag­ine is going to be the result in 10 years if this all works out?” And the answer is, well, my imag­i­na­tion ought not to be a gaug­ing fac­tor. If we in the end of ten years pro­duced every­thing I can imag­ine, we have failed. The idea is to pro­duce every­thing oth­er peo­ple can imagine.

I just hope that they will take the foun­da­tion that we give them – sor­ry about the pun – foun­da­tion in the Web. They will take it and they will build on it and they will be able to give us tools. Give the next gen­er­a­tion tools to be able to solve the huge issues that we have. Allow sci­en­tists to col­lab­o­rate togeth­er across the world to share their half-born ideas and be able to find the cures for dis­ease. And pur­sue those ideas about new forms of democ­ra­cy and mer­i­toc­ra­cy that had been cre­at­ed on the Web. Pursue them and maybe even in web sites, and by build­ing new web pro­to­cols, build new forms of democ­ra­cy that turn out to be actu­al­ly even bet­ter. Even more effi­cient, bet­ter forms of argu­ment, bet­ter trans­paren­cy, bet­ter account­abil­i­ty so that we can actu­al­ly move on and improve the way we man­age our­selves, OK?

They’re huge hopes and dreams in a way and it’s a big step. I’m very hon­ored and grate­ful that we’re mak­ing the first lit­tle steps towards the Web Foundation here tonight. Thank you.