#46: GeoCities and HellTubes
In a bygone phase the internet's history, probably around 1998, there was a website called “Three Kids Kicking the Crap Out of a Chair” (I know it couldn’t have been much later than 1998 because I was the correct age to find it extremely funny). The site, in ‘90s-internet fashion, simply displayed a bunch of photos of three kids destroying a chair in their backyard. Twenty years hence, I wouldn’t believe the site had existed outside of my imagination, except that I Googled it this week and actually found it, or at least found a link to it on someone’s Angelfire site that also hasn’t been touched in twenty years (‘90s Angelfire sites still exist somehow; Three Kids Kicking the Crap Out of a Chair sadly doesn’t). Everything I’ve just described is the stuff of a more innocent online age, a technological revolution in its youth, a fun, creative phase of the internet’s history for which we still wax nostalgic. Upon closer examination, though, you might notice that this preserved corner of the 1998 web bears uncomfortable similarities with some things we actually hate about the internet today. If you haven't guessed, I’m going to talk about YouTube.
James Bridle wrote a brilliant essay last year about the rise of weird, creepy, algorithmically-generated YouTube Kids videos. One lesson from Bridle’s piece is that context is important. With sufficient context, edgy children’s animation or footage of teenagers vandalizing their own furniture can be funny; without that context, it becomes slightly disturbing. A goofy, personal website with loud wallpaper and dumb GIFs situates teenage hijinks in a setting that makes sense. On YouTube, by contrast, everything seems a bit darker. YouTube has no context to offer—or rather, the context is just YouTube itself. Everyone criticizes YouTube now for its algorithms, which uncritically direct users toward the platform’s most malevolent content—conspiracy theories, violence, hate—but blaming the algorithms is too simple. Unsavory media has always existed in the world, and certainly always online, but it’s usually easier to avoid because you can tell when you’re approaching it: adjacent imagery on the website you’re scrolling through, the cover of a magazine, the people who are hanging around, the other side of the tracks. Context. On YouTube, where video is a commodity, everything looks the same until you’re already knee deep and it’s virtually too late not to watch.
We act like an “algorithmic” web is something new, a form of black magic that makes us do things against our will and turns our attention into money. This is all true in a sense. But before algorithms (whatever we imagine those to be), there were many other mechanisms for connecting us to things we didn’t plan on finding. In 1998, we visited GeoCities and Angelfire websites and clicked on things that looked funny or interesting, and later we'd maybe type URLs like “threekidskickingthecrapoutofachair.com” into a browser. If we ended up on websites with sinister, hateful imagery, we knew to leave. The true “algorithm,” then, is how we navigate these different environments. YouTube, too flat and featureless to navigate well, is a bad environment. Instead of complaining about obscure code we should ask for more of what we can all see: context.
I wrote an essay about blockchain's implications for cities on the Unchained Capital blog as part of their "Long Hold" series. I connected utopian modernism and JG Ballard's science fiction to the risks and promises of blockchain urbanism. It was awesome to write this because they gave me free rein and didn't require me to be a blockchain maximalist, although I am more of one after finishing the piece. Check it out.
Reads:
The perfect selfishness of mapping apps: Waze directions are optimal when only a few people follow them but make traffic worse when everyone does, which is increasingly the case.
"The Overwhelming Emotion of Hearing Toto's "Africa" Remixed to Sound Like It's Playing in an Empty Mall": Reintroducing a sense of physical space to music.
A weird neural network experiment from Robin Sloan that generates a gradient of intermediate sentences that "connect" two separate sentences that you enter. Fun to play around with.
Until next time,
Drew