A New Manhattan Project for Network Computing

Topic: A New Manhattan Project for Network Computing

Jonathan Zittrain: I want to solve the problem. I basically want to defend generative systems so that they stay at the core of our IT ecosystem. Manhattan Project, I think in retrospect, is an unfortunate phrase. What I meant by Manhattan Project was a collective effort, an effort with some urgency, and a sense of sacrifice, of actually being willing to take some time apart from our day to day to make a better system for everybody. What I fear it additionally conveys is a government project to the tune of billions of dollars, carried out in desert labs, in secrecy, that result in nuclear weaponry. None of that is part of the Manhattan Project that I'm calling for. So what I say is, it's time to realize that we can't naively assume these problems will solve themselves. They will in a sense. The market will provide us sterile alternatives, and then problem solved, no more generative problems. But also, no more generative benefits. I want to see if there is a way to realize a sense of community, wherever the generative problem was found, that may help to solve it. I mean, it sounds awfully naïve of me now, as soon as I say the word community, but you look at something like Wikipedia, which is undeniably, at this moment, useful. Who knows if it'll be around two years from now. It's entirely possible it won't. But Wikipedia is a collective hallucination. It is an ethos. It's a sense of, if you see a vandalized page, you should fix it. And out of the thousands of hits per minute that Wikipedia gets, the smallest fraction of people embrace the ethos or even know of it, but that's all it takes. As long as there are more people feeling enough identity with Wikipedia, who call themselves Wikipedians, and obsessively compulsively, instead of playing solitaire that week, or drinking beer, or watching TV, or having an avatar on Second Life, are choosing to sit on Wikipedia and argue about whether it's Kim Jong-Il, should he be the dear leader, or the great leader, depending on how he's deemed? Or whatever it is they're arguing about, they are improving this encyclopedia, and they are outnumbering, for the moment, the number of people who want to subvert it or vandalize it. I think that example of people undertaking something, not for money, but for love or for obsession or for fascination. That can be applied to other generative areas, like keeping the internet itself open.

Question: If you could speak to 5,000 web developers about your vision for their role in the future of the internet, what would you say?

Jonathan Zittrain: Well, there's a couple of things I would say. The first thing I would say to a group of web developers is, there are some communities we can form together that help us to solve some of the problems besieging the free web. So for example, there's an effort called StopBadware.org, that I've helped to start. It's at Harvard and Oxford Universities. It's funded by Google, Lenovo, Sun and HP. It has a couple of projects. One is to develop a piece of software that anybody can download that will relay the vital signs of that person's machine back to the rest of the herd. With it, you suddenly, for the first time-- I'm going to steal the phrase back from Bill Gates-- a digital nervous system, that lets you tell what the health is of the stuff attached to the internet right now. What's the weather like today? How many machines are actually rebooting hijacked, versus how many are perfectly happy? And you can measure that by the number of restarts they're doing, the number of popup ads per interval. There's some rough proxies for it. Even asking the user of the machine, "How happy are you with your machine right now?" It also means that before I run new code on the machine that might be dangerous, I could say, "How many machines out in the herd are running this code right now? How many machines, among people who have self-identified as expert web developers, are running this code?" and it will turn out that code like the Jessica Simpson screensaver, which comes bundled with spyware, is something that might be very popular in the rank and file, but not very popular among expert web developers. They have other vices than that. To be able to easily share that, and aggregate that data together, is an example to me of a community mechanism that can help people decide, according to their risk preference, what's good and what's bad, without having to turn to Peter Norton in a lab coat, or to the government, FTC banning stuff, to decide what's good and what's bad.

Recorded on: 3/8/08