THE FUTURE OF THE INTERNET
JONATHAN ZITTRAIN: It can be hard to talk about what we stand to lose if the internet and its surrounding technologies get a lot more rational, get a lot more cash-and-carry -- the way that so many other sectors of our economy that are thriving are -- but nonetheless, I think we have great risks at stake. And one of the best ways to understand it is to look back at some of the technologies that have been the most successful and ask how successful they might have been if there hadn't been an absence of gate-keeping that allowed them to thrive. So one example that may be a cautionary tale is something like Napster or peer-to-peer networking. That's the kind of thing that when a college undergraduate in Boston invented Napster and started sending the code around-- and the code was not that much rocket science. It was just, here's a way of putting files in a directory in your computer that you're willing to share with others. You are making them available for people to copy. It's like the essence of the internet from way back-- share some files, share some data. This of course focused on data that was music, files ending in MP3, and it completely undermined the prevailing business model of the music industry. You can see the music executives at the time were like, we don't get it. How is this allowed? Why isn't this being banned, it is surely illegal. And years of litigation still didn't substantially alter the fact that that kind of networking was greatly desired by a lot of people. They were willing to engage in it. Now again, this might be a cautionary tale. In a corporatised net that would never have gotten out of the gate. It also means that the kinds of pressures that brought about Spotify and other elements of the music industry, well they now see music as a service and something that you might pay a modest amount per month to have, or to get out of the kind of 15 US dollars per DVD or for a CD kind of land. Peer-to-peer networking just descriptively pushed things in that direction and in fact became popular because of the divide between what it really costs to move the music around and what the companies producing it wanted to try to gain. Other examples include things like Wikipedia. Wikipedia is one of the craziest ideas ever. Let's start with, I don't know, six articles that had been produced by academics getting paid by Jimbo Wales, the founder of Wikipedia, originally as something called Nupedia, and dissatisfied with the pace at which the academics were going, he and others experimented with having a wiki-- having the articles be in a form that people could come in and edit and make suggestions on. That was originally going to be the back room, but it turned out to be the live site where now anything can be edited at any time, more or less, run by volunteer editors who are themselves Wikipedians. This is the kind of thing that has been a smashing success-- hard to replicate. It's just so singularly successful. It's the kind of thing that I can guarantee you it would have made no sense if you presented it to Microsoft to say hey, instead of Encarta, what if everybody just edited everything? It's nuts. Or presenting it to CompuServe-- hey, why don't you create an area on your service where people can just edit stuff together and create an encyclopaedia -- it would never have flown. Here Wikipedia could gestate in a quiet corner of the internet with some devoted nerds writing articles about Star Trek, editing them, and then having the topics expand, having the number of participants expand, having the number of languages expand, and you now end up with the fourth or fifth most popular website in the world being this not-for-profit, collective hallucination set of information that is by no means perfect, but represents such an unusual way of using the technology to produce something that people find precious and valuable. That's thanks to the generative internet that we got that. Are there other examples? Sure. Anytime you see a business that got started as the proverbial two people in a garage, you're looking at something that they could go from conception to some kind of running site, up and live with people able to access it and then iterate it into new versions without having to negotiate to get access to those audiences. And many configurations -- partly driven by security, partly driven by other considerations -- of a future internet are ones in which that is not nearly as easy to do. And as soon as those barriers go up, you end up with ideas then that might get the way you shop a good idea for a television show to one of a handful of channels or producers, if they like it then maybe great they'll take it, because they are your gateway to that audience, and that greatly changes the nature of innovation.
THE GENERATIVITY OF THE INTERNET
JONATHAN ZITTRAIN: If you were to rewind time and play it back again, it's not clear you'd end up with the internet. I'm not sure you'd say the same thing about E equals MC squared, or truss bridges, or other pieces of technology that really are suited to what they do and are such the best solution, whether in theory or in practice, that you figure humanity would've gotten there one way or another eventually. With the internet, it reflects so many idiosyncratic choices about how to build a global network that it's important to remember they didn't have to be made that way. We could have ended up with a global network, or set of networks, that were a lot like the legacy telephone systems that prevailed from the early 20th century through into the early 21st. They might have been like CompuServe, AOL, Prodigy, The Source, MCI Mail. If you took a snapshot in 1982, '83, '85, everybody figured that the future of global networking was going to be something resembling those competing services, and one of them would just out-compete the others. Instead, out of left field came this unusual creature that we call the internet, and there's several things that made it unusual. The first of which is the folks who put its protocols together were not well-funded and had no particular expectation of making money from it. So in that sense, whatever you might know about Silicon Valley, the fundamentals of the internet were not as a dot com start up. They were instead as -- here are a number of ways that you could take existing networks and have them interoperate with one another, and that way you'd have a global network with no main menu, no CEO, no business plan, no particular committee even that runs it. A handful of functions, mostly what we call ministerial, merely administrative, are centralised, but the rest is pretty much what I'd call a collective hallucination. There's a commons that exists, in part through modest government subsidies to do this kind of research to produce the internet, but the rest is left up to the users to figure out what they want to use a network for. So in that sense, the essence of the internet is a set of protocols that allow any given points of presence that are connected to communicate with one another, and not have to worry so much about how the bits will work their way from point A to point Z. I have tended to call such technologies "generative technologies", and the core feature of a generative technology is that it tends to welcome contribution from nearly any quarter, and applied to the internet that means that in order to become a point of presence on the internet, to start exchanging bits with one other entity-- or perhaps if you think of yourself as a server, with hundreds of thousands or millions of other entities, if they are wanting to beat a path to your doorstep and get the bits that you have to offer -- that there's no gate-keeping to that. Anybody can set up on a kind of virtual hilltop a server-- maybe it's a web server, maybe it's some other kind of server-- and you're off to the races. That's very different from the CompuServe configuration of having to cut a deal with that company in order to be exposed to its subscribers. It's different from cable television. It's different from pay television or free television where there's a broadcast tower and a government will issue a licence or itself will be the broadcaster on unlimited bandwidth to an audience. And there's an interesting parallel kind of generative technology that really made the internet come into its own, and that is the typical end point. The thing that you would use to get onto the internet ended up being what we call the PC-- the personal computer. The PC's origins were less corporate and more hobbyist. Originally created by hobbyists for other hobbyists, and then by companies, but companies that in America would be called Heath or Heathkit-- kind of build your own clock rather than buy one. Not thought of as a big, broad market product, but rather a niche. That was the origins of the personal computer, and in 1977 when the Apple II was unveiled by Steve Jobs, 21 years old at the time, that was a computer that when it left the factory was not itself useful. You would plug it into a television set, turn it on, and you would be treated to a blinking cursor. That was all it was doing. It was waiting for you to write software, or to put in software written by somebody else that you've either purchased or been given or loaned, and that is another example then of a generative technology. Unlike say a smart information appliance, at the time -- late '70s, early '80s -- it might have been a word processor or something. You buy it, you turn it on, you get a word processor rather than a blinking cursor. You've got a document that you're ready to draft and then you could print it out. These PCs were general purpose, but originally to no purpose, and there was an understanding that anybody in the world could write code for it and circulate that code, and others, whether or not they were coders, could run the code. Two years after the introduction of the Apple II we saw Bob Frankston and Dan Bricklin of Boston Massachusetts produce VisiCalc, the first digital spreadsheet ever, and suddenly business around the world is noticing the personal computer. Now, it's what we would call an enterprise computer, because it's very useful in businesses to have spreadsheets. If you wanted to run VisiCalc you needed an Apple II. Apple IIs are flying off the shelves. Apple has no idea why -- they have to do market research to figure out what made their hobbyist computer so popular, and the answer was the generative nature meant people could code for it, didn't have to make a deal with Apple in order to get the code in front of people, you just needed your audience to have Apple computers, and sure enough they went out and got them so they could run something like VisiCalc. That generative technology meant that somebody actually from the southern cone, Robert Tattam, a researcher at the University of Tasmania in the psychology department could write something in 1995 called Trumpet Winsock (because he liked to play the trumpet), and that was kind of the keystone in the arch, the golden spike that allowed for the first time windows PCs to speak internet. So if you had a Windows PC and you ran Trumpet Winsock you could find yourself an internet service provider and get yourself online. And I think it's quite fitting that it was this gentleman's piece of freeware that actually was the gateway, rather than even that Microsoft had foreseen the importance to its own customers of the internet and had built internet connectivity into Windows, rather than building a paper clip that would tell you that it looked like you were trying to write a Word document, do you need help with that ... So thanks to PCs that could be repurposed and a network that was not a source of content, but merely a facilitator of its movement, we ended up by say, the year 2000, with this incredibly, doubly generative system that allowed anybody, anywhere to write code, to use this neutral network, to ship the code to others, to do it under any number of business models for money, not for money, for glory, not for glory, and that's how we saw the smart appliances fall by the wayside, and we saw the proprietary networks fall by the wayside. Now that's a snapshot as of 2000. In the intervening decade and a half we've seen a lot of growing pains. We've seen security threats to the internet. If anybody can write code, and it can easily work its way onto your machine how do you know the code is any good? And the answer is, you don't always know that. Microsoft, part of its way of talking about the importance of getting your code from accredited sources like Microsoft, likens it to a sandwich. If you found a sandwich on the street, would you pick it up and eat it? Probably not. So why would you do the same with your code? Now, the fact that you don't literally eat code is probably one of the answers. You can run it, and if there's a problem you could reboot the machine rather than have to go to the physician -- but that quirk has meant that as business models have come about to make it worth somebody's while to compromise your machine, to make it so that it answers to them far away rather than to you, the owner of the machine, we have seen great security problems arise. And my concern starting in around 2006, 2007, as I saw those problems on the horizon was that the cures for the problems might be as bad as the problems themselves in different ways. But we are unfortunately I think too often thinking that we are in a dilemma where we either have to suffer the vulnerability of arbitrary code running on our machine, scooping up our data, even worse, disturbing its integrity ... I don't know if it's worse to lose your spreadsheet or simply have cells within it transposed and you don't know until six months later that none of your numbers in your payroll make any sense, but that on the one hand-- to suffer those kinds of depredations-- either on your own machine or magnified to a merchant that has a server in which it keeps customer data that's off and running a PC operating system -- equally vulnerable. So that's on the one hand. On the other hand is locking stuff down so much that the wonderfully chaotic environment that gave rise to so much cool code out of odd and unusual corners could be stymied, that we could get back de facto to the worlds of CompuServe, Prodigy and AOL, because we will only trust code from a limited number of sources. Figuring out how to balance between those two undesirable end points to me is one of the fundamental questions about the future of the internet.
14. What is the purpose of the World-Wide-Web Consortium?
