This Facebook Network Visualizer mashes up the TouchGraph flash-vector-graphing software with Facebook's developer API and produces cool graphs like this one.

Real interesting to see how my network maps out. A giant tangle of extraordinarily networked and wired political/tech folks whom I know mainly from the past three years. And some outlyers, most of whom I've known for a longer time, but generally not as members of an interconnected posse.

Rich Orris has a special and odd role as the sole person whom I knew in college who is now a significant part of my more recent network. So while Rich is very much tied in to that central cluster, there are also a hard half dozen folks here with whom Rich is my only listed mutual acquaintance. I sort of knew this intuitively, but it's startling to see it visualized so clearly.

Of course the major factors confusing reality here are who's on Facebook and who isn't, and the lack of any contextual information about the strength or nature of my relationship with anyone. And yet, despite all this, the broad generalizations appear to be more or less accurate.

Footnote: Facebook is the first of the major social networking services I've tried (Friendster, Orkut, MySpace,) to offer the kind of serious developer API that makes stuff like this possible to create. That is really cool, and is a big part of why I expect FB to outlast the others.

Update: also interesting that while Facebook's structure is based around college-oriented networks, my college friends are probably the least networked individuals in the whole map. And yet the site always redirects me to "wesleyan.facebook.com". Yet another way community-driven tech strays from its intended purposes.

If you use facebook, take a screenshot of your map and send me the link!

This afternoon, an older relative of mine stopped by the house to coerce me into helping fix his company's website. Consequently, I have a viscerally renewed appreciation of the urgent need for intelligent web content management software.

He and his wife run a salon/spa in New Jersey. Their website was built a few years back by a frat-brother of his son-in-law. It's woefully out-of-date and awkwardly-constructed. Edits are made through a weird web-based file-management-and-wysiwyg thing, so just making basic basic adjustments is painfully tedious.

I am a little overwhelmed with this guy's willingness to jump headfirst into a completely bewildering medium. He listened intently and took several pages of notes while I explained things like the difference between editable text and fonted graphics, how to resize and replace an image, why changing a phrase's text-color to blue does not automatically turn that phrase into a link, how RGB hexcodes work, and why the page looks completely different in the wysiwyg editor than it does in the actual web-browser.

As we were finishing up, he summarized for me: "It's not easy for folks of my generation to understand this stuff. But what's frustrating for me is when I have instructions that say 'Do A then B then C, and then a window will pop up that says X,' but when I do those things no window pops up at all! Something completely different happens! And I do not have the intuitive understanding of things to know what to do next. I'm completely lost."

I had to correct him that, conventional wisdom aside, this is not a generational issue. If your instructions say something will happen, it ought to freaking happen, regardless of how old you are. And updating a basic ordinary brochure-style website should not require photoshop expertise or knowledge of RGB hexcodes.

Come on people, it's 2007, why haven't we got this problem licked yet?

Jon Chait has a real interesting cover article in this month's TNR about the NetRoots Movement. As with most mainstream political journalism, it's half studied objective history and half opinion and analysis, and the division between the two is completely unclear unless you're already familiar with the subject. Hillarious then that a large chunk of the article is dedicated to calling the NetRoots naive for their belief that all journalism is biased.

The major mistake he makes is to treat the internet like any other broadcast medium-- He basically paints Kos and Atrios as the Coulter and Limbaugh of the Left-- Stoller as Norquist-- Yglesias as, I don't know, Tucker Carlson? But anyone with a lower profile than theirs is irrelevant, unmentioned. After that, the article basically writes itself: Dean=Goldwater, EOM!

Obviously one way to criticise the article would be to blow-for-blow Limbaugh and Atrios and see who's more objectively repugnant. But the more important failure here is how Chait deliberately ignores the idea of decentralized autonomous online community, which is of course the single most important characteristic of the movement. You know, that thing that could just maybe make the NetRoots unique in all of political history. For Chait, blogs are basically mini-newspapers, not places for conversations. You'd think this would get a tiny mention somewhere in the article, right? No of course not, this is a Serious Publication, and such sillinesses do not merit mentioning.

Obviously the lockstepping tendency he's pointing to exists, and it's certainly something to be aware and afraid of. But this kind of "same shit, different day" approach ain't helpful or even accurate.

So Jen says to me, "Tim, I want to use your music for this NPR piece I'm working on." And I say to her, "You mean the one featuring brilliant instrumentalism from my good friend Ben Blum-Smith? Certainly you may use it, but it is completely muffled and unfinished, and your pieces are so smart and polished, why would you ever want to?" And she says to me "Well the piece is about the movie Big Rig, which is about truckers, and so your unique blend of banging breakbeats and country-twang banjos is perfect." And I say "OK fine, but I get to post your mp3 to my blog when you're done with it."

Note: This conversation may or may not have actually happened.

The final paragraph of this AP story is great:

Lawmakers also planned to press the Pentagon with questions still hovering over Tillman's shooting, including whether a Predator drone was flying overhead when Tillman was killed and whether it videotaped the incident. The military says no such videotape exists.

"Video-taped"? What is this, 1985?? The predator drone had what, a VCR inside with a VHS cassette? Beta-max, maybe? Thank you, Associated Press.

Client-work aside, I've launched two pet projects onto the web recently.

Oh-Ate.com is a real simple site launched in collaboration with Rich. It's an aggregator of the blogs, press releases, Flickr photos, and YouTube videos from the various presidential campaigns. It was inspired by a request from Zephyr, who wondered, is it really necessary to go to fourty-five different URLs to find out what all fifteen candidates are saying at any given moment? It also lead to some fun work with Micah and Josh on TechPresident.com, and Oh-Ate Version 2.0 is currently under construction in collaboration with Washingtonian.com's hardcore-looking webteam.

"NCLB?" was another quicky, built as an entry to The Sunlight Foundation's Mashup Congress Contest. This was a another collaboration, this time with Leah Nelson, a sometime journalist and longtime friend. To be honest, our hope was to find an obvious discrepancy between statewide educational performance and NCLB funding of each state, but in the end nothing so clearly unjust emerged. Instead, it's a complicated soup of data, and wound up creating more questions than it answered. Which is actually great, as it potentially opens the door to a more thorough and complex version in the future.

Both of these mash-up and consolidate data from around the web, but — more importantly — they begin to mash up the roles of "programmer" and "journalist" into something new and different. Adrian Holovaty, the Washington Post's lead developer, often speaks smart about this; how web-technology and database-wizardry will be as important tools for twenty-first-century journalism as photography and writing have been in the past.

This suggests one way of looking at the recent generation of web technology I haven't heard before; that the web is becoming a medium in and of itself. Up until recently, it's been conceived of only as a container for other media; words, photography, video, animations, etc. (There's the cliché that a bad website tries to be a television commercial or brochure or movie instead of being, you know, a web site.) But the new tools, the mashups et al, are a medium in and of themselves, one of course only in its infancy.

So: How will it transcend just being simple tools and clever gimmicks? In the coming decades, who will be the Annie Liebowitz of web-applications, the Seymour Hersh of mash-ups, the Ken Burns of data-analysis? What would that even look like?

M. Tim Jones, in the introduction to Artifical Intelligence Application Programming:

The problem is that technologies researched under the umbrella of AI become common once they're introduced into mainstream products and become standard tools. For example, building a machine that could understand human speech was once considered an AI task. Now the process that includes technologies such as neural networks and hidden Markov models is commonplace. It's no longer considered AI.

Rodney Brooks describes this as "the AI effect." Once an AI technology becomes utilized, it's no longer AI. For this reason, the AI acronym has also been coined "Almost Implemented," because once it's done it's no longer magic; it's just common practice.

...Products that integrate AI sell not because of their AI characteristics but because they solve a problem more efficiently than do products that use traditional methods.

Wired on Science Fiction:

You won't find the words "science fiction" in Random House's bio of Arthur C. Clarke Award-winning author China Miéville. Instead, he's called the "edgiest mythmaker of the day." Michel Gondry's The Science of Sleep? It's classified as comedy, drama, romance and fantasy, but not sci-fi, at Amazon.com.

Even Battlestar Galactica, the flagship show of (hello!) the Sci Fi Channel, keeps a distance. "It's fleshed-out reality," explains executive producer Ronald D. Moore in the sci-fi mag SFX. "It's not in the science-fiction genre."

...Chris Barsanti, a critic who dared to reference The Road in terms of sci-fi literature, said the phrase "science fiction" summons images of "space battles, aliens, mad scientists, time travel and the like ... fantasy with testosterone." So publishers, wary of putting their book into an "artistic ghetto," twist themselves into knots to avoid using the label.

So: Two ideas, both born in the 1930s, both perpetually weird. A kind of meta-weirdness that remains unexhaustably weird, even as it becomes increasingly intertwined with everyday life.

I wrote this Python script to convert AIF audio files into EPS vector-drawings.

Its utility is questionable, but leastways if you've ever desired an unusually high-definition graphic rendering of the amen break, now's your chance.

The script requires AIFC and PyX.

Herb Alpert & The Tijuana Brass- Summertime