Remote Viewing

An overly verbose, self-aggrandizing, bullshit attempt at a treatise on automated outsider art


For the last few months, I’ve been running two semi-anonymous projects that use unsecured network cameras. My goal was to present the appearance of these cameras as aesthetic, rather than intrusive; beautiful instead of filthy.

Insecam (now defunct) and ShodanHQ attempt to profit on the curiosities of casual voyeurs, piquing the curiosity of people that normally wouldn’t be a peeping tom. The mere shot of an interior space with people elicits many varying reactions, the details of which I’m not concerned with.

Once you see past the titillation of the intrusion, some of the vantage points have a composure of their own. The camera installer’s motives are unknown to me, but can be inferred. Sometimes it’s a public camera, used promotionally, so the scene feels composed and ideal. Other times the camera’s purpose is security, in which case the angles are harsher, the composition worse. The owner has crammed the device into some dusty corner, competing for space with cobwebs and condoms. But even the harshness of this viewpoint provides an aesthetic perspective. This view feels dystopian, and the hammer strikes a wire in ourselves and our personal impression of the world.

A large number of cameras were trained on construction sites. These will disappear at some point in the future as the crews complete their structures.

My favorite cameras function just to let the owner know that an object still exists and is functioning. I have a warm place in my heart for this strangely composed camera aimed at a telecommunications tower in Arizona. The sun’s angle crosses the field of vision of the camera during the afternoon. Its bright light overloads the sensor, creating a strange solar eclipse effect due to either equipment failure or a design flaw.

Most of the cameras I use I found via scanning sections of the internet that I was able to identify as belonging to business-class Internet service. Generally, these addresses are static, so that a camera won’t disappear within a few days or hours of me locating it. I’m not going to delve into the technical details of doing so. Many people with more network security credibility and knowledge have documented this sort of activity over the last few years. My main goal was I wanted unique perspectives, not the same creepy shit from those other sites.

@ffd8ffdb

My first project to incorporate insecure camera imagery is @ffd8ffdb. At an interval I select a camera from my list, and exclude cameras I’ve found that are in private spaces.

My script captures a frame, and gums it up with an Imagemagick script. I modify the colors in the YUV colorspace, crop out identifying information provided in the margins, and ensure the images are consistent. I use Wordnik to generate accompanying text and replace some characters with graphics characters. This is just for effect. @ffd8ffdb’s goal is superficial; I just like the way the tweets look. I enjoy that strangers find it unsettling, amusing, or even uninteresting. Like other Twitter bots, its unending tenacity is part of its charm. Many cameras go dark at night, most not having enough illumination to provide images. The bot doesn’t care and keeps stealing shots.

Treasure Column

My second camera bot is a YouTube bot, stealing video from a smaller list of cameras, mixing the colors down into grayscale. The frame rates of these cameras are variable, as fast as the network can deliver them, with some jitter. The end result is a rough time-lapse, and though each video is roughly a minute, the actual time elapsed varies.

The cameras are silent, so I’m using audio from a specific category on Archive.org, processed through SoX to create long ambient pieces trimmed to the video length. Most sound like a wire brush scraping inside of a pipe, but some have a pleasant, accidental melody that matches up with the imagery.

I arrived at the look of both bots by tweaking values in scripts for hours. There was no primary goal but to look the way that they do. There are no secret codes, no hidden messages, and they are not numbers stations.

What I learned

It’s hard to choose cameras that aren’t too transgressive. Every one of them is at least a little intrusive. I had to remove cameras from the last that showed something normal at first but later showed something a bit too borderline.

I think a signifying detail of good generative art is that it surprises even the owner. Even when the piece is complete, it stays unfinished. They either need maintenance or continue to generate unexpected content. I stopped looking at my bots’ content, but had several people pointing out strange tweets and videos.

When working on a content-generating bot, there is a point where what you intend for the project to be gives way to what the art wants to be.