Terra Media: How Sovereign Is Your Building Mr. Algorithm?
In every mediated act – in every form of expression – we always miscommunicate.
In the following investigation the question was posed: Could Machine Learning be used to train an algorithm to recognise degrees of ‘sovereignty’ in buildings?
In other words, could a technical analysis in machine learning reveal the ambiguities in meaning expressed by the appearances of things?
In mediation one should not seek the communication of absolutes. It is impossible. We only ever miscommunicate. In this study, the algorithm broke down all absolutes despite trying to recognise them. We have hacked away at the very boundaries drawn around meaning. Notions of identity, of façade semiotics, even of morality fall flat. Fragments of the world are neither in the territory of ‘tagged’ nor in ‘not-tagged.’ They are instead in a tertium quid, the terra media. Everything is thrown onto the line where the abstract nature of being is put into question. An object-in-mediation is nothing more than the line we draw around it. Communication is more than the exchange of data. Architecture is about designing and sharing experiences through miscommunication.
Below are some spreads from this investigation that culminated in the video which shows real-time sovereignty-detection in an urban context.
Terra Media is literally about the territory of the middle. Not of neutrality or objectivity, since the study on algorithmic recognition has already proven that impossible. Like any tool and language system there is always some degree of tainting reality; of miscommunication.
Terra Media is ultimately all about miscommunication. When we communicate we try to acquire knowledge in the act of naming. In art in the act of drawing we draw a line, a boundary which becomes the definition of what belongs inside the- line and outside-the-line. These two territories however are binary. Terra media is the territory on the line – the literal and conceptual line – between the knowns and unknowns. It is where the frame of the image-recognition algorithm tries to classify the objects it sees and what lies within becomes the approximation by percentage. The line is the site for where meaning (and the definition of the object) collapses. Essentially, fiction is always prevalent.
We only miscommunicate through stories and experiences. Even through the most empirical and mathematical of processes like the engineering of a machine, facts themselves are just as constructed. Given that the act of mediation is always a process of miscommunication why talk at all? There is more to mediation than just the exchange of data. Through repositioning our attitude towards the role of fiction any architecture is about the design of an experience. This study in algorithms was a study in meaning and its collapse and ultimate demise at the line.