The War Against Cocks

[My #BDF14 diary, written for Brandwatch, September 2014]

#bdfFTW – user submitted pixel designs in the windows of the Brandwatch building

Cocks, Hearts and Swastikas

This was the argument against from the outset. If we created a canvas and allowed anyone to post on it, entirely unmoderated, we were just inviting trouble.

It is part of the unwritten history of art that given any blank canvas, especially in an unfamiliar medium, the innate human instinct is to draw a penis. It’s what the male creative urge looks like; totem poles, tower blocks, rocket ships, sports cars and power tools. Phalluses are the default. I’m sure Picasso scrawled many in his time. And Renoir. Matisse. Hogarth. Freud. Dali definitely did. The renaissance painters quickly covered theirs with cherubs, to avoid going to hell. As an artistic tradition it goes back to prehistory, the earliest cave paintings. Genitalia first, bison, stick men and sky gods later.

the Cerne Abbas Giant

a Sim City user creation

From what I can tell, this is a predominantly male obsession. The female equivalent is hearts. Give a schoolgirl a blank sheet of paper and she will fill it with love. Give the same paper to a schoolboy and he will fill it with male members. This pretty much says everything you’ll ever need to know about the biology of gender.

If we were to build a digital canvas, and open it up to the public, might we simply be building the latest, most technologically sophisticated, gallery space for crudely drawn cocks? Or swastikas? Messages of racism / misogyny / hate? Should we expect the worst of humanity, or might we give them a little faith and trust them to use the platform in more creative and sophisticated ways, as was intended?

This was the experiment.

The Concept – Pong, Kong, and the Hello Wall

I first got involved with the idea back in March. Katja Garrood, Brandwatch‘s Creative Director, wanted to do something with the BW building in celebration of the 2014 Brighton Digital Festival, which the company sponsors. What, exactly, they weren’t quite sure at that point.

The building overlooks Brighton Station so, while it is a little awkward for a gathering, it enjoys high footfall and is to many visitors their first impression of Brighton. We had an opportunity to give these visitors a bright, digital welcome. Possibly in the form of crudely drawn cocks.

the Brandwatch building, in relation to the station

I looked into building projection, but there wasn’t really anywhere to project from within a reasonable distance. I also looked into doing something with lasers (frikkin lasers!), which was exciting, but suffered the same problem. But my first and favourite idea was LEDs. LEDs are cheap, very cheap, especially compared to the cost of a few nights projector hire. And I really liked the idea of doing something lo-fi and reusable. It meant we might even be able to scrape a little extra cash from the xmas lights budget too.

early concept sketch

I originally thought of it as a series of LED strips that we could hang down from the windows. The company was across two floors of the building so we could, feasibly, have mounted two strips under the second and fourth floors, giving us a broken rectangle to play with.

But time, logistics, and fear of the landlords trimmed the concept down slightly. To do it safely we couldn’t just dangle stuff out of the window, we’d have needed proper fixings on the outside of the building, and to close the road outside while we deployed the cherry-picker required to fit them.

This would have added a lot to the cost, as well as making us rather unpopular with Brighton’s commuters. So, after some consultancy, the concept moved indoors. We’d do it in the windows rather than against the brick, meaning weatherproofing, seagulls, and the possibility of injuring passing pedestrians were no longer an issue.

The second question was interactivity. We wanted to do something a bit more than just a display. It had to engage the audience in some way. This was 2014 after all, when passive consumption was something only your mom did. But our bold new Century hadn’t yet decided a de facto standard for how one interacts with a building, so we needed to come up with something clever.

I did a few camera-based experiments, tracking bodies to see if there might be anything interesting in the patterns of the passing commuters. But, post-Snowdon, this passive surveillance angle didn’t feel quite right somehow. Face recognition and analysis were ruled out too for similar reasons. The idea of a building that frowned back at you if you were looking grumpy tickled me, but I wasn’t sure I’d want to be that commuter, returning to Brighton after a long day, having the piss taken out of them by a building.

camera tracking demos

Another idea was a game. Around this time I’d been doing another BW skunkworks project, experimenting with multi-player networked games. I’d built an MMPG, a Massively Multiplayer Pong Game, purely as an exercise in data gathering, so my aesthetical thinking was very much in the area of retro games. “25 years of the web” was one of the themes of that years festival, so while the ultra-lo-fi graphical style of Pong was overshooting by about 10-15 years, it was still looking in the right direction.

perpetual pong visualisations

The tiered nature of the building was screaming Donkey Kong at me. I really wanted to turn the building into a playable Donkey Kong game, multiple Marios leaping barrels to reach the top. This was our lead idea until we moved the concept indoors.

donkey kong on the TRS-80

The biggest problem with the game idea though was that it would require a clever networking solution. We’d need a wifi hub covering the outdoor area if we wanted decent real-time interaction. But I knew the BW network guys were maxed out, so I didn’t want to have to rely on them. I’m not sure what they could possibly have been doing that might have taken priority over the building of a tower-block-sized 80s video game, but, whatever it was, I wasn’t going to disturb them.

So it was a third idea that got the most traction, which was to use Twitter as the interface. I’d seen Hellicar & Lewis’s 2010 “Hello Wall” project in London; an interactive billboard at which you tweeted instructions – e.g. “more blue”, “circles slower” – to influence the display. I loved the elegance of this interaction, sophisticated enough to be engaging, simple enough to be autonomous and self-managing.

Hellicar & Lewis’s “Hello Wall”

It was very hard to draw a penis on the Hello Wall. Although people tried, of course. If anyone did manage to pull it off, I thought, they’d have earned their moment.

So our concept was aiming at that same sweet spot, between satisfactory interaction and porn-free abstraction. We could keep it reasonably cock-proof by way of the lo-fi aesthetic. We’d allow people to draw on our windows, giving them full control of the pixels they put up there, but if we only provided the mechanism to target one window pane at a time they’d have just an 8×16 pixel grid in which to express themselves. They could draw whatever they wanted, but it would be so lo-res it would take considerable skill to create a design that was unambiguously rude.

We hoped this would be a sufficient prophylactic for the more sensitive eyes of Brighton. If someone were to complain of filth, any submitted design would be so abstract the interpretation could only be in the eye of the beholder. If they possessed a brain that interpreted certain pixelated blocks as something other than a rocket ship, a finger, or a chimney, it was their fault for bringing their own perverted pareidolia to our innocent, community-minded art project.

As you might imagine, we had a lot of fun at concept stage, as at that point there really was no limit to what we could do. But suddenly it was May, and September, the month of the festival, was fast approaching. It was time for reality to come a’knocking. I had to work out how to build it.

The Build

One of the joys of having the resources of a large company behind you is that one can have a stupid idea, with only a rough conception of what might be involved in realising it, and still stand a good chance of finding someone in the corporate mass who might understand what you’re talking about. I sent the following email to all Brandwatch staff in early May:

Hi Team Brandwatch,

Do we have any electronics hackers on this list? 

We have a potential project that may call for some dirty home-brew ‘tronics, 
to do something really cool. It'd be helpful if there was anyone in the office
(or any of the offices) who has played with this stuff before - and might like 
to play with it some more on work time.

If you know what the words "arduino", "pi", "piezo" and "breadboard" mean, 
you're the kind of person I'm thinking of.

Ping me a mail for more info…

From this I hooked up with Ben Slaughter and Jimmy Taylor, who had considerably more electronics experience than me. I could make lights flash on a breadboard, but my only plan for scaling it up at that stage was just to do the same but, er, bigger. But Ben and Jimmy knew about power distribution, what type of LEDs might work best, where to source control boards, and all that stuff. They pretty much carried the project from there.

By the end of the month Ben had built our first prototype, while I fathomed out the software side. It then took two further prototypes to work out how we were going to scale it. Once we had our third (built within a fixed wooden frame) we were rolling. We actually knew what we were doing. Jimmy then took care of speccing out the full build and ordering the tools and components we’d need to start a production line.

Our first three prototypes

Having spent so long splashing around in the concept and prototyping phases, we were left with a shrinking amount of time to build the full-sized panels. We had four weeks, and day jobs, so the big question was how many panels we could feasibly construct in that time.

But again we had the advantage of 1) a large staff of people, 2) the support of Katja and the rest of the BW management, who were 100% behind our crazy idea. With the help of a few more mass emails, a Doodle poll, and some pleading we recruited some more bodies, Brandwatch’s finest, who willingly volunteered their downtime and lunch-hours to do a bit of soldering, wiring and woodworking. All in the name of our shared quest to see some pixelated genitalia on the side of the building we worked in.

The production line constructing the full-sized panels

With the build progressing nicely, I could relax a little and turn my attention to the control software and the interface. I found myself getting seriously anal about text encoding, spending way too much of my time working out ways of getting the maximum amount of visual information in a single tweet (tip: four byte kanji encoding!).

The interface at

It was work I eventually threw out though, in the name of simplicity. We decided to allow people to send just a single image, or three frames of animation, each frame encoded as a 32 character hex string. The interface would construct their tweet for them. This was enough. Anything else would have been unnecessarily complicated.

our respect wall

There were 29 names written on the wall by the end of the month. Twenty-nine enthusiastic volunteers well deserving of respectful nods and fist-bumps.

But there was also credit due to the other 70-80 staff in the Brighton office, who good-heartedly stifled any complaints over us taking over their breakout area for a month, carefully stepping over the soldering iron cords and fishing stray LEDs out of their sandwiches. The build was a truly collaborative project, making the finished piece every bit the product of the company, not just a handful of us.

The Launch

We made our deadline. We completed 14 panels, and had them mounted, networked and working for the 1st Sept, the opening of the festival.

the mounted panels, from inside and out

For the first week it displayed only ambient animations and pre-defined text tickers. We didn’t want to open up the Twitter interactive element until we had a way of monitoring the display remotely, i.e. a webcam pointed at the building. This actually proved to be one of the most technically demanding parts of the project.

We couldn’t have done it without the utter loveliness of the staff of Grand Central, the pub over the road. They very kindly allowed us to leave a laptop and webcam in an upstairs room, trained at the building. But, with them being so accommodating, we couldn’t really then hog their broadband just to stream our image. So we had to work out a way of getting an internet connection over there.

Luke Whiting, a late addition to our unofficial BW skunkworks team, rode to the rescue here, coming up with a way of extending the office wifi to a building 200 metres away by way of a tight-beam transmitter. Which to me was akin to witchcraft. But it meant we could then broadcast the pub livestream in full 1080 HD. Technology!

Luke’s tight beam hookup

So, once it was visible to the world (not just Brighton) we could publicise the interface. It had taken a week to get the livestream working, so it was Friday 5th September, five days into the festival, that we went fully live. I let Marketing know. I tweeted the link. I sent an email round. Then I went out and got pissed.

Now, before you begin to question my professional standards, let me give a little context. Drinking on an empty stomach is not my usual launch strategy. And is not something I’d recommend. But this wasn’t just negligence, there was other stuff going on. Please allow me to explain.

The Digital Festival is a big event in the Brighton calendar, and that first week had been a heavy one. It had been a week of long days and late nights; conferencing, networking and socializing with the various interesting folk the festival bought to town.

The night we went live was no exception, the big event of that evening being the opening of the New Sublime digital art show at Brighton’s Phoenix gallery. This was a show I had co-curated and so never considered, even for a second, skipping the Private View to shepherd the launch of my other big project that month. And the getting drunk part was kinda expected. Because this is what you do at Private Views. Everyone knows that.

The boards had already been up there for a week, displaying placeholder content, so it was only the user submitted content we had turned on that day. We had flicked the switch at lunchtime, in daylight, when the pixels were invisible, and everything had gone fine. Which may have been why I was perhaps a little overly-nonchalant as the sun went down.

It wasn’t until I got home that I first checked the video feed and saw my first set of user submitted creations. They were the work of one author, @brightonstation:


Go Home, You’re Drunk

My first reaction was to laugh. Hard.

First I was happy it was being used. Second I was pleased that the constraints were being hacked already. I had wanted to, subtly, expose the method by which the encoded tweets were constructed, hinting at a greater control one might have if one learned how to sidestep the interface. But, even in such a tech-savvy town as Brighton, I never expected it to be hacked quite so quickly. It may seem odd to some, but this was why my first feeling upon seeing a profane message in large letters on the side of our building was one of community pride.

Third, I laughed because it was actually kinda funny. It just was. But when I saw the image being retweeted, and realised it may have been up there for a few hours, and no-one else was interacting to over-write it, I started to get a little nervous. As did Katja. And Giles, Brandwatch’s CEO. And Will (CMO), over in the states. And the rest of the American marketing team, who had to explain via email that “twat” was a much stronger swearword in the US than it was over here. Something, in all my years, I had never been aware of before.

Then there was a rather tense half-hour when the police showed up. I had Katja on text-message, Giles and the US team on email, and a police van on the webcam, parked in the bus depot watching our display. Thanks to the livestream we’d set up earlier that day we could all watch from afar, in glorious HD, as a group of officers debarked, marched up to the entrance of the (then empty) building, and peered into the darkened lobby, trying to work out what the fuck was going on.

We hastily posted some smileys to remove the message, and the police eventually got bored and drove off. Off to deal with some riot, murder, burglary, domestic, or whatever, hopefully something that might be considered a greater threat to the peace than a brightly lit swearword on an office block.

Meanwhile I hastily (drunk) coded a bit of PHP to allow the US team, who were at that moment in a more sensible timezone, to be able to see who had posted what and so could block people if necessary. Something I’m pleased to report we haven’t, as yet, had to do.

I didn’t think I was going to get any sleep that night. The remote access I’d hastily set up in the day wasn’t working, so I had no way of turning it off. The only way I could censor anything, if required, was to tweet new designs. But the family-friendliness of the submissions improved greatly into the early hours, and when I saw @SimplicityCom‘s “Go Home You’re Drunk” around 1.30am, I took it as a personal message. I was already home, and I’d sobered up quite a bit by then, but the sentiment convinced me my LEDs were in safe hands. I could succumb to my tiredness and drift off.

“go home ur drunk”, by @SimplicityCom

I left the display in the hands of the people for the evening. The next morning I awoke early and checked it before the sun came up. I was greeted by a row of hearts. The female artistic urge, not the male, had won the night. Love had prevailed.

Katja’s grab of the sunrise with hearts

The Unwinnable War

Saturday gave me time to do some fixes, whilst Katja dealt with the press. IT helped sort out the remote access, so I could code in a shutdown time and we could all hopefully sleep easier. The only other problem from thereon was reminding the cleaners to turn off the lights. And, of course, the inevitable stream of cocks. Shapes which became burnt into my retinas over the weeks that followed.

It was a war we couldn’t win. And so didn’t fight. The best censorship method was simply to post something else to the same windowpane, with maybe a gentle warning via twitter for persistent offenders. But it never really came to that. There were a few, but it never got particularly bad.

The fact that every design had to come from a Twitter account may have helped, encouraging self-censorship. Every cock had to be signed by the artist. Or maybe it was just that the users who engaged had more maturity than we’d given them credit for.

Anyway, it wasn’t just cocks. Or hearts. Or swastikas. There was gold in there too. I built a gallery to explore the submissions, in animated-GIF form, and saw declarations of love, abstract art, clever multi-panel stuff, dancing stick men, agit-prop, ghosts, skulls, and lots of smiley faces. On the whole, the forces of good were winning.

The installation is up for another few days (at the time of writing). So if you want to send us some pixels, and add to the gallery, get in now while you can. It is coded to switch off at midnight Thursday 25th Sept.

And if you must obey the urge, if you really must draw a penis, please think of the children, and make it an ambiguous one.

-- 23rd September 2014 --

« »
Generative Art book cover Generative Art: A Practical Guide
(Manning 2011)

The definitive guide to the philosophy and practice of coding as an artistic medium.

Foreword by Marius Watz.
Novelty Waves book cover Novelty Waves: A Short Book About Digital Art
(Leanpub 2014)

Five long articles and two short stories exploring digital art.

Foreword by Jeff Noon.

Name your own price.