The cybercandidates don't necessarily have polish or glamour. They are not telegenic. Dean often looks physically awkward, and Steve Forbes was a hopeless nerd in front of the camera, but netizens see them as geeks like us. The awkwardness comes across as a marker of authenticity, especially when backed by the candor and take-no-prisoners attitude people associate with such candidates.

If radio turned out to be a more intimate medium than old style stump speeches, the Internet is turning out to be an even more informal medium, encouraging candidates to show us what goes on behind the scenes. Political talk shows like Inside Politics on CNN and dramas like The West Wing have educated voters about the work that goes into constructing a campaign's image and themes. The Dean campaign takes this fascination with "spin" one step further-allowing voters to collaborate in shaping the image of our political leaders. It is no accident that many of the best ideas about how to use the Web in the Dean campaign have come from unsolicited e-mails or that the Dean Web site's most popular feature is a blog written by his staffers about their day-to-day activities.

The telegenic candidates had to be cool under pressure. Never let them see you sweat, as perspiration-drenched Nixon learned in 1960. But many of the cybercandidates-Buchanan, Perot, Nader, Dean-have portrayed themselves as mad as hell. When Dean says that he represents the "Democratic wing of the Democratic party" and accuses his opponents of not standing up to Bush, he shows the spirit of the veteran flame warrior who blasts away his opposition. Pundits question whether such candidates have the temperament for the office (a charge directed first at McCain and now at Dean), but supporters give more money every time the cybercandidates shoot off their mouths. Television is a push medium-political messages come to you whether you want them or not. The Web is a pull medium-you have to actively seek out information. So far, the hot-blooded candidates are more effective catalysts in drawing supporters to the Web than those who come across as calm and collected on the tube.

Once they come to a site, though, netizens demand data. Ventura's campaign offered much more detailed position papers than either of his major party opponents. The telegenic candidate depends on sound bites; the cybercandidate, on posts. Part of Dean's appeal is his willingness to let us in on his thinking process, to show us step by step the logic that shapes his positions. Not all Americans will want all of the details and may not read through all of those position papers, but the blogging community can help customize the message to the concerns and attention spans of specific constituencies. The key thing is giving the impression that we can drill down as deep as we want and still get more information. Yet this focus on his thinking process leaves Dean vulnerable to traditional media, where provisional ideas get treated as definitive positions.

Will the Internet improve U.S. democracy? It will certainly change it-and for the moment, it is certainly enlivening the early days of the 2004 campaign. What we need to watch for is how the traits that add up to a successful cyber-campaign play out on broadcast and cable television, which operate on very different paradigms. So far, the traits that make for an effective cyber-insurgency don't necessarily play well on the evening news, and we may be seeing the cybercandidates getting caught in the crosscurrents between the two very different media.