Façade, Interactive Drama, and Expressive Artificial Intelligence

By Charles Khosla | November 25, 2024

By the end of the twentieth century, digital technology had begun to integrate with the very fabric of cinema, theater, literature, and nearly every form of art.  The emergence of digital media as an artistic apparatus led many critics and academics to fearfully anticipate a complete overhaul of art itself; in the new millennium, artwork no longer had to be tangible to qualify its existence.[1]  Yet despite the analog art community’s technophobia, it is important to remember, as Peter Weibel once suggested, that human vision has always required “machine-assist[ance],” meaning that digital technology can be another instrument in an artist’s toolbox rather than an existential threat to the creative industry (Youngblood 28).  Michael Mateas’ and Andrew Stern’s interactive game Façade (2005) illustrates how digital technology can create new forms of storytelling, translating Weibel’s notion of machine-assistance to the world of video games.  A one act domestic drama in the vein of Who’s Afraid of Virginia Woolf? (1962), Façade was one of the first prominent games to incorporate text-based generative artificial intelligence within its playthrough to create immersive experiences for its players, rendering anthropoid kinesthesia into the digital world.  By implementing AI into its coding, Façade provides players digital bodies that live within Mateas’ and Stern’s simulated reality.

Figure 1 – When playing Façade, users can talk to Trip and Grace by typing their dialogue, which the duo will respond to accordingly based upon the underlying AI coding.

Façade’s revolutionary text-based engine offers users more agency during their playthroughs than most games, granting users greater freedom within the fabricated space.  Mateas’ and Stern’s narrative centers upon the disintegrating marriage of Trip (Andy Bayiate) and Grace (Chloe Johnston); gamers assume the role of a friend visiting the couple for an evening in their New York City apartment.  Upon starting Façade, users can choose the name of the protagonist which Trip and Grace will address them as and speak to the pair by typing their dialogue onto their keyboard to initiate or trigger new conversations based upon the AI encoding (figure 1).  In their 2005 essay “Writing Façade,” Mateas and Stern describe this textual machine-assisted gameplay as a channel of “discourse acts,” as gamers can interact with Trip and Grace through their own speech (6).  Façade’s discourse act gaming system allows players to have sovereignty over the virtual environment; with this text-based dialogue, users could choose to emotionally mediate in Trip’s and Grace’s relationship, aggravate their anger towards each other, or even flirt with the duo.  Through this AI engine, Façade permits its users to experience “global agency,” where the discourse acts between the user and the digital couple can influence future conversation topics and the story’s ending (“Writing Façade” 4).  Whilst most video games engage within some level of agency, Façade demonstrates how the player’s individual actions, no matter how small, can realistically reshape or redefine the entire narrative.  Based upon the gamer’s textual inputs, Façade can end with Trip and Grace splitting up, reconciling their differences, or even evicting the user from their home if they behave inappropriately (figure 2).  Façade’s global agency not only encourages users to conduct multiple playthroughs to experience all these different possible conclusions but conjures a virtual universe that grants gamers the same autonomy of reality.  The AI engine provides its users a digital mouth, allowing them to speak freely within Mateas’ and Stern’s digital realm, rather than picking from a list of preselected choices (the latter being standard interactive story model behind a Choose Your Own Adventure-novel).  While Façade’s AI engine is admittedly crude by modern standards, with certain inputs can be misinterpreted or ignored by the system, its innovative textual gameplay introduces the sense of speech into the medium of video games.

Figure 2 – In this playthrough of Façade, a user was immediately dislodged from the apartment after revealing Grace’s secret affair to Trip.

With its Weibelian design, Façade’s AI engine recontextualizes its users’ existence in the digital mise-en-scène, amplifying the boundaries of the human consciousness.  In her book Interactive Cinema (2024), Marina Hassapopoulou cites Façade as an example of expressive AI, a system where computational algorithms mimic anthropoid behavior to “alleviate the existential fear that humans will be replaced by machines;” rather than threaten or undermine the status quo of mankind, expressive AI attempts to conciliate the differences the natural and mechanical worlds (206).  In his essay “Expressive AI: A Hybrid Art and Science Practice” (2001), Mateas further elaborates upon the term, describing it as a communicative mediator between authors and their audiences and an outlet for new modes of expression and human cognition (150).  While expressive AI does not disguise its mechanical nature, it seeks to translate the behavioral manner of human beings into the digital realm.  Through Façade’s use of discourse acts, gamers almost always undergo an entirely different experience through each playthrough, triggering distinct narrative beats based upon their textual inputs.  Façade’s algorithmic nature may be inherently artificial, yet players can suspend their disbelief because the expressive AI system realistically replicates the social patter and mannerisms of a splintering married couple.  Although Façade’s coding and language processing tools mean there is a certain limit to the possible gameplays, Mateas’ and Stern’s design reconfigures the player’s position their digital world, in effect transforming them from bystanders into Mateas’ and Stern’s fellow programmers, with their coding contributions being their spoken textual input.  As Hassapopoulou suggests, Façade cloaks its technological artificiality by allowing users to “cocreat[e]” the story with expressive AI coding (150).  Mateas even argues that through expressive AI projects like Façade, audiences could not only have pleasurable experiences during this algorithmic communication but even learn something about themselves (“Interactive Drama, Art and Artificial Intelligence” ii).  Expressive AI is more than just a gaming device, it can potentially provide therapeutic, educational, and emotional support for its users.  Perhaps if expressive AI can make machines become human, then humanity can finally embrace a mechanical creation as its emotional and intellectual equal.

Façade’s employment of expressive AI completely reshapes the relationship between audiences and artists by gifting gamers the capability to talk within its virtual reality.  Juxtaposed to literature, movies, and even most video games, Façade refuses to constrain its players within a small set of predetermined options; Mateas’ and Stern’s interactive drama allows users to express themselves freely as the digital protagonist.  Façade demonstrates how digital machinery can further enhance and expand the human consciousness into the world of art; spectators are no longer just voyeurs of a fabricated space but active participants in this fantastical universe.

Works Cited

Albee, Edward. Who’s Afraid of Virginia Woolf? 1962.

Façade.  Directed by Michael Mateas and Andrew Stern, Procedural Arts, 2004. Windows/Mac OS X game.

Mateas, Michael, and Andrew Stern. “Procedural Authorship: A Case-Study of the Interactive Drama Façade.” UC Santa Cruz, UC Santa Cruz, 2005, https://users.soe.ucsc.edu/~michaelm/publications/mateas-second-person-2007.pdf.

Mateas, Michael. “Expressive AI: A Hybrid Art and Science Practice.” Leonardo, vol. 34, no. 2, Apr. 2001, pp. 147–153.

Mateas, Michael. “Interactive Drama, Art and Artificial Intelligence.” Interactive Drama, Art and Artificial Intelligence, Carnegie Mellon University, School of Computer Science, Carnegie Mellon University, 2002, pp. 1–273.

Rodowick, D. N. The Virtual Life of Film. Harvard University Press, 2007.

Youngblood, Gene. “Cinema and the Code,” in “Computer Art in Context: SIGGRAPH ’89 Art Show Catalog,” Leonardo, supplemental issue, vol. 2, 1989, pp. 27–30.


[1] In his book The Virtual Life of Film (2007), D.N. Rodowick went one step further, claimed the digital technology “wants to change the world,” reshaping the public’s perception through a never-ending pool of information (174).  The digital form has completely transformed the human consciousness; people now often demand the Internet for their mental processing.