A large number of current „social movement prediction models“ are automated using Natural Language Processing (NLP) methods. Poetry presents probably the greatest challenge for this computational way of processing texts in natural language. It has a high density of ambiguity and usually plays by its own rules.
In my doctoral project I am developing a concept i name „Adversarial Poetry“, a subversive resistance practice to compose politically motivated texts in a way that they are misinterpreted by common NLP prediction models.
Not just governments and their secret services have an interest in surveillance through recording and analyzing the structures and individual actors of social movements. In evaluating their social media accounts, also private companies and platforms like Twitter or Facebook created their own instruments of domination to automatically detect „abnormalities“ and pass them on to the relevant authorities in natural language.
As a result, activists need to create new spaces to communicate on the web under the radar of surveillance. Therefore, they have to level self-determined and self-organized platforms for participation within socio-technical spaces of action. Spaces, which are always based on text since computers are semiotic machines.
Literary currents and movements of the last century were, so to speak, the forerunners in a development of alternative forms of language that are very difficult to read for computers today. In connection with the social dynamic of hacktivism, strategies can be developed that provide activists with possibilities to play with contemporary instruments of domination.
It is possible – if one knows which rules to break and how to break them without bringing down the overall set of rules – to destabilize this socio-technical space of action. How one appropriates this space with poietic rule-breaking will be the challenge of the project.