May 15, 2008
[b@10] Jonathan Zittrain
[
After Dean Kagan (correctly) identifies JZ with the Berkman Center, JZ begins by talking about the importance of the fact that the Net at its start was unconstrained by a need to make money. They therefore didn’t have to count how many people were on it or how much they were using it. So, they made it so anyone could get on just be hooking in. Anyone can build on it. It explains the hourglass shape of the Net: Diverse media, diverse tasks, all going through Internet protocol. [Live blogging: JZ is a great speaker, and my scribbled notes don’t come close to capturing the flow, much less the texture, of his talk.][Note: I’m posting this without re-reading or spellechecking so I can go join the hallway chat.]
IP reflects the constraint and lack of constraint. Ethernet, for example, relies on social conventions to keep the hardware following the protocol. Same with email. The natural way to create email would have been to set up an authenticated database. Instead, email assumes a distributed email address and assumes people won’t spoof addresses. Solutions to problems generally are postponed until the problems arise. He points to the informal, unpretentiousness of the IETF as typical of the founding attitude of the Net.
JZ points to the “generative” power of having programmable computers attached to an open network.
The heart of his argument: The social conventions may not be enough to prevent the “turning of the barge” of the Net. We are losing the fight. Our PC’s assume that codes running there are good and desired, whereas PC’s (and Macs) frequently run programs we don’t understand or want. Vint Cerf has said that perhaps 250M machines around the world are running code waiting for commands to do something evil, e.g., botnets. “This is an absurd situation. We would not allow our cars to be used for joyriding.”
E.g., Pakistan YouTube by exploiting a network weakness: An ISP altered its routing tables so that other routers thought it was one hop from YouTube, so YouTube traffic went that way.
“The Internet is a collective hallucination that works so long as we don’t stare at it too carefully.”
He puts up a 2×2: hierarchy <> polyarchy, and bottom-up<>top-down. (Polyarchy means lots of people can try out lots of things.) JZ says bottom-up = generative and top-down = sterile. He puts the Internet in the bottom-up, polyarchy quadrant. “This area works great until it doesn’t.”
So, then what do you about it? Have the government fix it? A Patriot Act for cyberspace, as Lessig fears? Sometimes the government does intervene, JZ says. E.g., it adjudicates domain name disputes. But he shows the ITU’s architecture for fixing the Internet: an enormously complex chart that is the opposite of the simple Internet hourglass. The changes is to quality of service (not best effort) and fully compliant with all regulatory requirements. That’s the future Internet being designed for us. Likewise, PCs are being locked down. He points to the Mac that says Macs are better because all the pieces come from a single vendor. Even with the coming of the iPhone SDK, all apps have to go through the Apple Store, subject to Apple’s regulations, as announced by Jobs: Nothing illegal, malicious, privacy violative, porn, bandwidth hog or “unforeseen.” JZ thinks that this type of lockdown, which is coming, is the worst of both worlds. Locked down appliances and gated communities.
He points to FaceBook being used as an alternative email, but with pretty hideous terms.
So what do we do? Perhaps the lower left quadrant: bottom-up hierarchy: Bottom up solutions with enough adherents that it has the heft of a hierarchy. We need social buy-in. E.g., Wikipedia has technology that makes the price of mistakes not to high and that lets people talk and come to agreement. That’s the kind of buy-in we want for stopbadware.com at the Berkman Center. But this also requires that we import some practices from the regulatory sphere, e.g., an appeals process to get people off the stopbadware list. We need ways of expressing our social preferences with the condience that they will be respected. But, he says, some of these social techniques can be subverted. And we don’t always agree on norms. When we don’t find solutions in that quadrant, we find ourselves turning to government.
Q: Scottt Bradnor: JZ is sort of right, but hyperbolic.
Q: David Reed: I sort of agree. There were a lot of skeptics when the American democracy was founded. The problems we’re suffering through are not the iconic ones. They’re more subtle. They’re about power shifts and whether commercial entities are really helping us. I’m not convinced the problems are insurmountable.
