Twin Speaks is an artificial duo in dialog. Development started in 2018. The title Twin Speaks is referring to the cult tv-serial Twin Peaks by director David Lynch where pervasive feeling of magic plays a dominant role. Dialog is always a combination of movements and voices. Both coming from the body and - for the living beings - melted into one. Dialog is an abstraction on the level of interpersonal relation, where some fusion happens between the two parties caught in a dialog. Movement is complemented by a counter-movement and voice by a counter-voice. Of course, the voices can also melt into one song. Or turn into incoherent babble.

Twin Speaks is firstly an interactive-generative structure, taking on the metaphors of identities and roles that emerge within the process of communication.  Computer vision further allows for the objects to sense the human face and follow it (“personification of the subject from the part of machine”) – and opens the space of potential communication between the objects and the visitors.

The two entities are connected via wireless, so they they both have some idea about the dialog.

When an entity turns to a visitor: this is a look. They can look at each other and prove they are separate entities. If in-synchro they melt into one – the two eyes. When one follows the visitor and the other looks at the first - this is the view of the third (“the other”) - on the situation. There is quite some variations possible and therefore a good study of identities and roles that these entities can take on themselves – visitors included. Of course it is a metaphor of human relations that as an artist I was always interested in.

The autonomy of artificial creatures used to be quite magical when noticed by the visitor. It causes the personification of the entity from the part of human observer. Entity-object is recognized as entity with identity. Next step would be giving it a name (“a house pet”). The existing study is not finished: it uses some random algorithmics - jumps between basic movement positions to form variety of movements – stochastic methods and more coherent moments of (a)syncronized patterns. Artificial intelligence (AI) would probably enhance the concept of this kind of autonomy – with more meaningful interactive-generative results. But at some point it would become an everyday thing - the fascination, magic would be gone. This moment may already be now.