Wednesday, April 16, 2008

Jonathan Zittrain: The Future of the Internet - and How to Stop It (review)


Author: Jonathan Zittrain.
Title: The Future of the Internet, and How to Stop It.
Publication: New Haven: Yale University Press, 2008. ISBN 978-0-300-12487-3, hardcover, 342 pages, endnotes, heavily indexed, 3 Parts, 9 Chapters.

Well, the cover is certainly entertaining. A railroad track junction, almost from a model railroad, splits,with one track leading over an escarpment into what looks like the Great Rift Valley in Africa, the other staying on the trail. The visual metaphor is clear enough.

The rest of the book is not cinematic; in fine print, and a somewhat thick and speculative writing style, it would make for a difficult book report. But Zittrain has an important message: the public has come to perceive the Internet as intrinsically “unsafe” and companies, partly out of liability concerns and partly because “covered wagon days” business models run their course, might be inclined to lock down a lot of the “generativity” that has enabled user-generated content and added an important element of both competitive efficiency and pseudo-direct democracy in Western society (not so much so when misused). Already, as with the iPhone “fiasco,” there is a tendency to “re-tether” users to specific vendors.

“Generativity” is a rather intuitive concept. I guess I’ll say the visitor should buy the book to see his exact definition, on page 70.

The book has three parts: (1) The Rise and Stall of the Cooperative Net (2) After the Stall (3) Solutions.

In the first part, he gives a pretty informative history of computing culture, particularly going back to the rise of commercial mainframes in the 60s where everyone had to “get IBM” (including programmers). There was, for a long time, a conceptual distinction between organization (government, company), and the individual / family / microbusiness. When personal computers were introduced in the late seventies, there was at first a concept that they were more like “appliances.” It wasn’t long, though, before computers could manage most of the “business” functions within a home. I recall that even in the early 1980s, when I owned a Radio Shack TRS-80 (the Tandy Center decorated downtown Fort Worth when I lived in Dallas), it was apparent that the home PC and even rudimentary printers (like the dot matrix –ices and then the daisy letter writers) could revolutionize life for freelance writers. Even so, in the early days, magazines like Writers Digest were recommending computers “specialized” for writing. (in 1979, a Lanier office word processing system still cost $15000). Some computer brands (like Atari or Commodore) seemed designed more for the gaming and entertainment market. (Remember the Osborne?) In computer courses at community colleges, instructors talked about “general purpose computers” (IBM and its competitors, all of whom eventually failed in the iron horse market). By the mid 80s, home PCs were becoming much more “general purpose” and even home laser printing was becoming practical. (I remember a predecessor of both Word and Word Perfect: Q&A, which was pretty effective). The “appliance” concept still held, though. Consumers had to choose between VHS and Beta formats for VCRs (we know what happened -- or do we, as the remnants of the litigation over VHS still show up today in copyright cases), and soon were teased by CD’s which would “last forever” (especially for classical music hobbyists who used to wonder about tone arms and elliptical styli to play vulnerable vinyl recordings). To “get published” or get a movie made, you had to go through well established industry agenting channels, without exception.

There were various early experiments at home online connectivity – Compuserve, and FidoNet, in the 80s, while “the Internet” was perceived as the province of academia and government and perhaps big business. Naturally, networking operations at first had a lot of control over what got “published.” In the early 1990s, toward the end of the first Bush administration, the National Science Foundation eased the Internet into public access. A particular connectivity tool within Microsoft (Winsocket) would become critical to generativity. Home computer hobbyists learned how to set up their own servers (sometimes on 386 machines, even) to connect to the Internet, while “average Joes” depended on ISP’s like AOL and Prodigy (and Compuserve, still, and then Earthlink), which first tried to work profitably on a proprietary content and subscription model (especially for email). In those days you could function at 2400 baud on dialup with a PS-2 (486). With Windows 95 (and winsocket), the idea of more generalized protocols (http, ftp, gopher, etc) took hold, and the “generativity” of http (particularly) was apparent. Companies like AOL started offering personal publishing around 1996 (with “hometown AOL” and an embedded FTP, which was a little clunky to use). Microsoft offered Front Page, with the ability to connect directly to a domain and edit online, easily (although “Front Page Extensions” were tricky to manage, all now replaced with “Expression Web”). By around 1997 or so, it was practical for an average person to set up his own domain and use a package like WS-FTP (or Front Page) to edit content. Self-publishing of text documents in large volume, capable or reaching the entire planet, was at hand. By 1998, it was apparent that search engines like Google really would give “average Joes” global reach. Zittrain points out that by the end of the century, the traditional publishing and media industries were becoming worried about how they would “compete” with no-cost amateurs who wanted to be noticed and would offer often valuable content for free or at low prices. Even so, some concepts (like e-books and Softlock) expressed the old “appliance” model.

Probably the first big concern, though, was security. Viruses had been known since the Morris affair in 1988, and at first they were spread mostly with diskettes, and then email attachments. Eventually they became more sophisticated until eventually even connecting to the Internet at all incurred some risk. Zittrain sketches out some scenarios, of how a total cyber meltdown could conceivably occur or, more likely, some individual users could find all their files trashed and their machines rendered useless without rebuilding, even when they thought they had kept up anti-virus protection. Much more likely, though, is gradually infection of machines with spyware, or possibly having machines taken over as zombies with DOS attacks (a problem widely reported in the media in 2001). The vulnerability of Microsoft operating systems (along with their omnipotence) led to the massive “service pack” update around the end of 2003 and the need for most home users to have broadband to keep their machines adequately protected by both frequent anti-virus updates and Microsoft security fixes.

Home users, and operators of websites for small business or just personal political expression do encounter potentially serious risks. ISPs could cut them off if their macines become zombied, although it’s not always clear that ISPs have an incentive to do so, or to help customers after some security disaster has occurred. Hence, the risk of kneejerk. Websites, even those of small businesses or individuals, could be hacked, either randomly (most likely) or because of an “enemy” issue, and then quarantined by site advisor services or even search engines. It’s difficult for individuals or small businesses to “appeal” these actions by other private parties, partly because the “locking” parties tend not to have much financial incentive to do much, and because their employees (even in abuse departments) may lack the skills to do much. “Acceptable use policies” of ISPs and various business facilitators usually include legal indemnification policies; these are rarely enforced or pursued (with the notable exception of proven spammers) for rather transparent business reasons. In rare and perhaps elaborate cases (maybe the stuff of John Grisham novels), a particular individual could be “framed” by a hacker for committing some particular computer crime. In practice, however, actual wrongful prosecutions have been rare. Auto and property insurance companies can try to cover these with “umbrella policies” but must contemplate the actuarially unknown risks of dealing with “amateurs” on a public stage, and even questions like what makes someone an “entertainer.” Media perils companies have generally not been willing to insure “amateur” bloggers (at least those dealing controversial political stuff like LGBT issues) on the Net because of the unknowability of risk.

At some point, the discussion of generativity (vs. lockdown and “tethering”) comes back to social engineering and social values. Zittrain covers much of this with his detailed discussion of how Wikipedia works and enforces its ethical standards. (He refers to its founder as “Jimbo Wales”, whom other writers characterize as a “curiously powerful man”). Zittrain spends a chapter on the “reputation” problem, although not so much from the viewpoint of the harm to employment prospects from videos of underage drinking (what Dr. Phil calls “Internet mistakes”) as something much broader and of a real concern to business. As I noted in the previous review, young entrepreneur Cameron Johnson talks about a business person as equivalent to his reputation and brand, as if they were synonyms. Zittrain feels the same way. He suggests the idea of “reputation bankruptcy,” as if various private mechanisms will gradually rate people’s “reputations” the way Fair Isaacs calculates FICO scores. It’s not clear how a reputation can be cleaned, however, other than through the moral suasion efforts of companies like “Reputation Defender.” (I covered Dr. Daniel Solove ‘s book on online reputation on this blog Jan. 12). Perhaps there will evolve new “aging” standards about how long references to “non-celebrities” can remain on totally public websites.

Zittrain points out, as have many others, that with the generative Internet, the a different and somewhat amorphous idea of “privacy” has evolved. In small town America, no life was private; in urban America, life could be anonymous, an asset for the LGBT community dealing with throwing off McCarthyism. With technology, at first there was an emphasis on the value of anonymous speech and access (which EFF and the ACLU always vigorously defend), now the desire for “fame” and “pride” has complicated the picture. People want to express themselves and be known for what they can say, as on the Internet, but they don’t want the National Security Agency eavesdropping on their conversations. (Zittrain covers the FISA and Patriot Act issue briefly.)

There could be some benefit to biometric identification of users in some cases. There needs to be careful thought to the responsibilities of ISPs and content facilitators, even though it seems important to protect them from downstream liability (as with Section 230 of a 1996 telecommunications law), even though some "pro child protection" see that exemption as controversial. The Internet makes defining precise legal standards for child protection (from both explicit and "implicit" content) problematic, as we know from litigation (COPA). On spam, he sounded optimistic about Sender-ID, and discussed how captcha boxes (and attempts to defeat them) work.

I think that the self-expression has even more social significance than Zittrain or Solove state. Self-expression goes along with “individual sovereignty” and the idea that one can be the absolute god of his own social connections. This lives in a certain tension with “family values” and ideas of blood loyalty that used to impose some socialization on everyone.

In his Conclusion, Zittrain discusses Nicholas Negroponte ‘s initiative “One Laptop Per Child” with special laptops called XO ‘s with a design that encourages a kind of “localized generativity.”

Correlated post on Zittrain's position on Network Neutrality, here.

No comments: