H&F: How did you design the facade?

DS: We took a photo of the building from the square, 200 meters away. Then we started to manipulate this feature. We wrote a small piece of software that would just move lines inside the image. You get these codec errors, it’s very common. Like when you upload an image and there are some bytes missing.

CW: The glitches we got in the beginning made the colors get really bright or go away totally. So we played around a bit looking for a glitch or an error that would give us an aesthetic transformation of the facade. What we were looking for was an error that would be buildable. When you start glitching floor plans it becomes impossible.

DS: Then we went to an architect to talk about possibilities.

CW: There were still some worries that the building would stay up.

H&F: So when you took the design to the architect, you didn’t know if the building could stand?

CW: Actually we already thought that the [existing] columns were not necessary to keep up the building, but of course, we’re not sure, we don’t read architectural plans every day. [The architects] immediately saw that the idea would work. but there were some things they had to figure out, like, what will happen with the water from the roof? Because our design cut the water spouts.

People don’t usually look at a building going “How does this work?” But yesterday at the opening, people were talking to each other, saying “Where does the water go now?” People started thinking about this structure that they take for granted normally.

H&F: How does this project fit in with your previous work, which is largely conceptual and performance-based?

DS: It relies very much on the piece we did before. When we started this project, Random Darknet Bot was still running. There we deal with algorithms that go and do things by themselves. They produce glitches, errors. We don’t know how they work anymore. The idea was around us because of that project.

CW: The error has a long tradition in art, not only as a visual concept but as something that can explain underlying algorithms. What made this work really something we wanted to do was this transformation of something that’s very short lived into a structure that’s totally permanent.