Literary Machines: Between Drama and Artificial Intelligence

comment 1
actualization / artificial intelligence / behavior / being / cinema / cybernetics / drama / existence / image / interactivity / interface / language / one / reverie / story / theory

Insofar as it tends towards a studied negation (or counter-actualization) of our numb existence, endlessly dramatizing the existential narrative of escape, contemporary cinema conveys an intense and disturbing truth about modern reality. It is not just the easy and everyday dissociation that life is a movie, but more surprisingly, that movies have become indistinguishable from our real lives. Drama is not merely a terrifying absence, simply the dissolution of the synchrony of the One, but an enigma even more puzzling still– the cinematographic interface is already an incontrovertible diachrony, a renegade communication across an abyss of broken myth. Cinema: the art of time turned in upon itself. The surface story splits the world in Two, by drawing a sacred circle (or rectangle) in which larval intensities can escape from their segmented order of time (or space.) Images transcend history.

Drama, in its fiery and stormy heart, is the story of the becoming-One, the becoming-Image, the imaginary One whose becoming exceeds the “legal” or “logical” boundaries of the story world. Above all the telling draws together the one who can answer a profound question, that is, the one who tells responds to a specific call. While a matter may hold terrifying secrets, and suddenly seem to overflow the encircled space with dark and terrible visions of other worlds, nonetheless, like a vampire, the storyteller would not have come without being asked inside. Who knows, it may even be that the real story cannot even begin without a pre-original apology, a gift or other crude offering of gratitude, on the part of the interlocutor. For drama is not recitation, it is spoken reverie.

Only original impressions are magical. Only if God were a playwright, could existence be so profoundly miserable and magnificent at once; and in such a delicate symmetry! Who must be convinced the universe expresses themes which are by turn unfathomable or tragic or transfiguring? We feel becoming as dramatization, already “actualization.” Now what is needed is to digitize creativity, to automate the playwright, to quantify writing. For becoming is the body upon which the author, the artist, the cinematographer play, these mad directors of active and reactive forces of intensity. All elemental and narrative forces are available at their palette. The complete artist — who is thus in absolute hell, for he knows a little about horizons. Now it is revealed: God writes straight in crooked lines, or put another way: being is indifferent.

And becoming is a black hole. Yes, the movie is our life. Why deny it? It is already over and the story hasn’t even begun. The medium itself is fluid, full of noise. Yet, despite its transience, cinema is still richly traversed by pure elemental forms. There are sublime moments of crystallization in cinema that rival or exceed anything I have ever experienced.

For the space of the drama, the screen functions as a turbulent transparence.


Preface (Emergence from Noise)

This study seeks to explore the relation between dramatic theory and artificial intelligence. The goal is to shed light on a broader and more difficult question: how do we reconcile authorship with autonomy? In terms of this deeper question of creativity, it will suffice for the purposes to investigate only peripherally the question of human expressivity. Indeed, almost our entire concern will be with a-linguistic feedback systems, and in particular, systems of non-human communication. The first goal of this study will be to develop a theory of “autonomous” interfaces. This model is then applied to several particular examples of interactive systems, followed by a description of the computational as well as phenomelogical features of ‘deeply generic’ interactivity.

Consider the problem of dynamically generating a dramatic story. Since playable characters would be to some degree autonomous to the plot, how do we achieve dynamic interweaving with the plot? The future parts of the story must be actively rewritten, the world must be warped about the active characters, entities in the game-universe controlled from somewhere outside. The space of user behavior is precisely the critical gap of narrative control between the plot and the characters.

In terms of developing interactive drama, we are still awaiting some legitimate cybernetic interconnection across disciplines. In short, we have need for generic representations of plots and characters and then rules of mixing them up, arranging and transforming them to respond to user interaction. In terms of design, then, we need to create a machinic interface between the author and the story space. Even though the fields are abstract, a real engineering is required here. We need to build literary machines which can resolve this break between plot and autonomous characters.

Efforts similar to our study have in the past produced conceptions of drama-managers (micro-agents who arrange plot-atoms combinatorically to conform to certain desired dramatic functions.) Eventually most studies have concluded that the problem lies in the authorial interface. You don’t want to have to program a thousand different variables to describe a character, but on the other hand, you don’t want very few variables. In fact, you want a natural interface which develops a full character based on a partial description, with a lot of creative guessing and inventing in between. That is, you want to be able to describe in natural language what a character is going to act like.

This behavioral dimension of activity is difficult to characterize; certain languages (like ABL) even make behaviors the elemental basis of programs. Again, we run into similar problems, if almost all the drama and narrative is being generated by the plot subsystem, how do you account for the uniqueness of new users? How do we dynamically generate characters without simply scripting behaviors? The answer is to have many layers of evaluation, and more complex levels for more complex interweavings characters and plots. The key is pragmatic variation: it is a tool which can be used at the lowest level of language generation as well as at the highest level of overall dramatic management.

Let us briefly anticipate our results in order that the situation be made clear. There is an inescapable gap in the philosophical and computational discourse on interactivity, the explicit question of bridging this gap together narratively. Now, in many ways, the problem of natural language understanding is strictly beyond the scope of this study, though obviously the theoretical apparatus we develop is intended to be generic enough to effectively model certain interesting features of human languages and stories. Examples of such models will be sketched out in the final section.

(sorry, more on this soon)

The Author

mostly noise and glare

1 Comment

  1. Two things, to start with I suppose: you mention quantifying writing…That’s in the introdution:rhizome to ATP:

    “D+G insists upon ‘Stratometers, deleometers, BwO units of convergence. Not only do these constitute a quantification of writing, but they define writing as always the measure of something else. Writing has nothing to do with signifying. It has to do with surveying, mapping, even realms that are yet to come (4-5).”

    That’s straight from the beginning of my notes…

    Anyway, there’s that aspect of it, and there’s the fact that literary theory can really help approach a lot of the problems, ‘frame’ them, so to speak. In fact, I’m tempted to get that metalepsis book back just for this project that you’re developing. The idea of metaleptic narrative spaces could provide means of modeling the comparative topological structural folding of narratives, with feedback loops disparating the series, blah blah blah… you know Logic of Sense could really kick ass here too…

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.