Creative Destruction: A Pedagogical Process with the FrankenProject

Calvin Olsen

#Creative Destruction: #A Pedagogical Process with the FrankenProject ##       by Calvin Olsen, <br>             Kelsey Dufresne, <br>             & David Rieder ##Welcome to [[FrankenProject.]]

#It's alive! #          [[It's alive!]]

Our discipline needs to examine both kinds of writing. In addition to examining writing as "the thing," meaning final products that may be entirely or even partially comprised of alphabetic text, we need to investigate the various kinds of writing that occur around—and surround—writing-as-the-thing. —Jody Shipka, //A Composition Made Whole// Being touched, in the sense of being moved emotionally, signals the very real experience of something immaterial pressing itself on our emotional state and changing it, and the capacity for us to touch someone or something else in a similar way. —Aubrey Anable, //Playing with Feelings: Video Games and Affect// #[[Table of Contents]]

#[[About]] #[[Introduction]] #[[Processes]] ##   [[The Pedagogy of Critical Making]] ##   [[Triangulating Transductions]] ##   [[On Learning and the Mediated Glitch]] #[[Conclusion]] #[[How to make your own FrankenProject]] #[[Glossary]] #[[References]] #[[Sway]]

#Introduction Mary Shelley's 1818 novel, //Frankenstein//, has had a lasting impact on both European and U.S. popular cultures. Initially drafted in 1816 during a rain-soaked and stormy "Year without a Summer" (Friedman & Kavey, 2016, p. 1) in the Swiss countryside near Lake Geneva, the British gothic novel has proven to be endlessly adaptable, serving as the inspiration for novels, plays, films, and hypertext stories. In an introductory video for Morgan Library and Museum's exhibition in celebration of the 200th anniversary of Shelley's story, John Bidwell states, (align:"<==")+(box:"===XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX=======")[It's so capable of interpretation. You can do it in so many different ways. You can make high literature. You can make popular culture. People have parodied it. They've abridged it. They've imitated it. Practically any medium, //Frankenstein// has been successful. (The Morgan Library & Museum, 2018)] For the authors of this web text, two doctoral students and a faculty member in the Communication, Rhetoric, and Digital Media PhD program at NC State University, Shelley's novel was the impetus for our [[Collaboration]]. When we agreed to work together on a new digital interactive project just before the start of the 2020-2021 academic year—the fated first year of the COVID-19 pandemic—we took particular interest in Victor Frankenstein's methodological use of [[Creative Destruction]], or the process of breaking down and destroying in order to construct something new or destroying while simultaneously creating. This method is seen in blackout/erasure poetry, in a phoenix burning to be reborn, and even in how Victor Frankenstein destroys and destructs numerous bodies to rebuild one anew—or how he creates a monster while destroying himself. Recognizing that our experiences collaborating from a distance represented a type of sutured inventional process—one that can be described in terms related to pedagogical goals—we began taking note of the stitching that held us together over the months as well as how developing a [[Physical Computing]] project in pieces aligned us with both Victor and the monster. Therefore, we identify that we have enacted vicarious methods that speak to the text, as well as the broader fields of digital humanities and rhetorics, [[Critical Making]], [[Transductive Rhetoric(s)]], [[Physical Computing]], and [[Discursive Design]]. With this in mind, we collaborated on a touch-based, digital interactive project that would connect its users to some of these themes. Related to the epigraph from Aubrey Anable's book, //Playing with Feelings// (2018), we had hope that a touch-based project would lead to an engagement with the various themes we were exploring. We found that this project served as the vehicle through which we aligned Frankenstein with our own monstrous [[Creative Destruction]]. As such, we were able to inhabit Victor Frankenstein by going through the experiential-based process, "reanimating" Shelley's novel in a digital-interactive context. The nonlinear process that we went through echoes the themes and tensions of Shelley's novel and emphasizes an interdisciplinary and open-ended pedagogical approach and method to reconsider and reevaluate literature through critical and digital means. Thus, we explore the role of the glitch and [[Glitch Art]] in connection to Mary Shelley's work, Victor Frankenstein's collage-style construction of a body from dissonant beings, and our project's processes, all of which strived to bring together disfamiliar pieces—all resulting in a conglomerate of modalities, experiences, and lessons taken from a set of parts that worked together in unforeseen ways once combined. Our inventional process, our distal, collaborative experiences during the pandemic of COVID-19 working on the project, and most of every aspect of the project itself resemble(d) this kind of dynamic assemblage of parts made whole in a process that renders wholes parts as the glitch inherent in a hyperlinked life is enacted. For the three of us, there was pedagogical value in much of what we worked through this past year. For that reason, this webtext is ultimately meant to contribute to theories of digital pedagogical invention. Our sutured-together experience as distal practitioners working on a [[Physical Computing]] project represent something transferrable to other projects, and perhaps, too, the classroom. We took on a project based around themes of fractured identity and the contradictory dynamic of [[Creative Destruction]], and our separate but overlapping experiences and the uneven, glitchy process resulted in a fascinating byproduct of a hyperlinked life. What we discovered ourselves, surrounded by an inventional process, is that which is not really found in the final "thing"—but rather a process that echoes the practical and adaptable type of intelligence epitomized by metis as well as of a material process involving Styrofoam, conductive paint, 22 gauge wire, sensors, bolts, glue, and 3D-printed body parts. It's a process that seemed caught in a strange, creatively-destructive engagement with the materials, practices, and bio-socio-cultural contexts in which we were working at a distance this past year. [[Processes]] [[Table of Contents]]

#About <iframe width="560" height="315" src="https://www.youtube.com/embed/I_InbGEfgv4?si=MD8DSVgS1-_KaQq1; title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe> FrankenProject (or the end result of FrankenProject, at least) is a touch interface that remediates—or, facilitates user-generated glitch readings of—Mary Shelley's //Frankenstein//. It is made up of three main components, which we produced and developed over the course of a full calendar year: 1. A Styrofoam head (with bolts!) and multiple 3D-printed hands, all of which are covered in conductive paint and wired to an Arduino 2. A transductive engine programmed to "know" when someone is touching the head or hands and translate the user's physical touch into data 3. A [[GPT-2]] poetry generator that creates a poem on screen and layers other audiovisual elements on top of it based on user input Working together, these three components can respond to users. Every interaction with FrankenProject is at least slightly different, because when you touch the head or (at least) one of the hands (align:"<==")+(box:"===XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX=======")[1. the human body "pulls" energy from the object you are touching. 2. that object registers the amount of voltage lost to transduction (how much the current changes based on how/how long/where you touch). 3. the program (written in Processing with a Python grandfather) performs a [[Mediated Reading (n.)]] of //Frankenstein// based on that loss. 4. a new, glitchy reading appears on a screen, along with a number of surprises.] The audio tracks in FrankenProject are based on an array of seven .mp3 and .wav files, which were extracted from YouTube videos and then cut down and optimized in Audacity. While FrankenProject is running, two or more audio clips are playing on a loop simultaneously, and their amplitudes (volume levels) change based on user interaction. A timer is used to swap out audio clips as the project progresses. The swaps or changes usually coincide with changes to the layered background images. As you, the user, interact with FrankenProject, you are literally and figuratively connected to Shelley's novel. In particular, you join Victor Frankenstein on his journey from interested party to creator to spectator as his/your creature/assemblage acts in/on the world. The rest of this article will break down our creation process and offer steps for making your own FrankenProject, which we hope you'll share with us. FrankenProject currently lives in the Ricks Annex at North Carolina State University, home of NC State's program in Communication, Rhetoric, and Digital Media. Its audience is everyone, and we hope it sends you back to the marvel that is //Frankenstein//. It has produced more than one audible gasp from audiences, so its current future (driven by the promise of more gasps) is to exist as an installation/experiential art piece at NCSU Libraries. #Calvin Olsen #Kelsey Dufresne #David Rieder [[Introduction]] [[Table of Contents]]

This is a presentation of FrankenProject in slide format built using Microsoft Sway. If it's glitchy (as it's sure to be eventually), <a href="https://www.calvin-olsen.com/frankenproject-sway.html" target="_blank">you can download a PDF from Calvin's website (at least until that, too, becomes glitchy).</a> <iframe width="760px" height="500px" src="https://sway.cloud.microsoft/s/TiZeCIrmTjOcTIoR/embed" frameborder="0" marginheight="0" marginwidth="0" max-width="100%" sandbox="allow-forms allow-modals allow-orientation-lock allow-popups allow-same-origin allow-scripts" scrolling="no" style="border: none; max-width: 100%; max-height: 100vh" allowfullscreen mozallowfullscreen msallowfullscreen webkitallowfullscreen></iframe>

"Creative Destruction," by Olsen et al. by KairosEditors

[[Table of Contents]]

#Processes (Authors' note: If you've already read through a process, you can skip this contextual information by scrolling directly to the bottom and choosing a new Process.) Beginning in the summer of 2020, we met over Zoom biweekly to develop and create the FrankenProject. We utilized Python, GPT2, Natural Language Processing, Processing, Arduino, and much more to create this project—simultaneously learning how to operate these applications, tools, and languages as the project developed. Throughout the process(es), we centered our creativity, work, and product on a glitch aesthetic that prioritized defamiliarization and [[Creative Destruction]]—emphasizing a visual creative destruction that aligns with Victor Frankenstein's creation and destruction of the human form (and of himself) with his monster. All the more, we centered on the role of touch. In Shelley's novel, touch serves as the force through which Victor Frankenstein creates and brings to life his monster, but it is also the method that the monster utilizes to kill those around him. Throughout the Covid-19 pandemic (and especially in its early days), touch has been as feared as much as it is desired: We sanitiz(ed) all surfaces, refrain(ed) from touching and being near others, and miss(ed) human interaction as we practice(d) social distancing. We used Mary Shelley's text as a springboard from which we could explore numerous technological avenues and pursuits, focusing our work on creation and experiential-based learning—while simultaneously constrained and inspired by the consequences of COVID-19—working collaboratively and continuously remotely, online, and in solitude. It's important to note that we recognize the immense privilege of being able to do so—of being able to continue our work, and being able to work, at a time in which so many face(d) discontinued work and increased dangerous and unsafe working conditions. We also recognize that we were able to foster a community through this continued [[Collaboration]]—despite never being in the same room or place, or even never meeting face to face. We recognize that many around the world are operating in similar isolation and/or feeling deep solitude—craving connection with family, friends, and even strangers. Through this, we have found a deeper sense of sympathy for the Frankenstein's monster and his experience watching a family from a distance for years, longing for companionship and community. Thus, at a time when we crave touch while simultaneously fearing it, we were inspired to center touch in our undertakings. Each of the following sections explores our personalized and individual learning experiences throughout this project, as well as the processes we employed to explore [[Creative Destruction]] through the glitch. While each team member took charge of specific aspects and components of the project (and this webtext), we continually taught each other and worked in tandem to produce the FrankenProject. ##[[The Pedagogy of Critical Making]] (Led by Calvin) ##[[Triangulating Transductions]] (Led by David) ##[[On Learning and the Mediated Glitch]] (Led by Kelsey) [[Table of Contents]]

#Conclusion We worked together throughout this project, but we also worked independently in silos of semi-separate components of the project—ultimately striving to stitch FrankenProject together in a unified and cohesive whole. The experience of digital humanities-centered [[Critical Making]] provided each contributor the opportunity to build new relationships with technologies and coding during a time in which human touch was as restricted, guarded, and limited as possible. As such, this project served as the vehicle through which we aligned //Frankenstein// with our own monstrous [[Creative Destruction]]. Doing so allowed us to inhabit Victor Frankenstein by going through the experiential-based process, remediating (and "reanimating") Shelley's novel in a digital-interactive context. The nonlinear process that we went through echoes the themes and tensions of Shelley's novel and emphasizes an interdisciplinary and open-ended pedagogical approach and method to reconsider and reevaluate literature through critical and digital means. This project is firmly centered in process-based learning—where failure, error codes, dead-ends, and "I don't know how to do this yet, but I will watch tutorials on YouTube" sentiments were not only common occurrences but part of the overall experience of learning, creating, and destroying (though not always in that order). As our world still faces the COVID-19 pandemic, we are actively interested to see how further research will continue to explore the role of mobilities and touch with digital humanities and [[Critical Making]] spaces and capacities. [[References]] [[Table of Contents]]

FrankenProject centers critical engagement with touch and is based in part on a series of unique readings of Mary Shelley's novel //Frankenstein//. We feel that Shelley's novel is a compelling text for exploring touch today—as well as themes like science, creation, identity, and the consequences of the polarization of thought and feeling. ##[[For teachers and classroom implementation]] ##[[Discussion questions]] ##[[Table of Contents]]

#References Anable, Aubrey. (2018). //Playing with feelings: Video games and affect//. University of Minnesota Press. Bak, Meredith A. (2018). Taming monstrous play: STEAM learning, maker culture, and monster-making media for children. //Comunicazioni Sociali: 2//, pp. 218–231. Balsamo, Anne. (2011). //Designing culture: The technological imagination at work//. Duke University Press. Boyle, Casey. (2018). //Rhetoric as a posthuman practice//. The Ohio State University Press. Boyle, Casey, Brown Jr., James J., & Ceraso, Steph. (2018). The digital: Rhetoric behind and beyond the screen. //Rhetoric Society Quarterly, 48//(3), pp. 251–259. (link: "https://doi.org/10.1080/02773945.2018.1454187")[(goto-url: 'https://doi.org/10.1080/02773945.2018.1454187')] de Bruin-Molé, Megen. (2021). Monster theory 2.0: Remix, the digital humanities, and the limits of transgression. In Eduardo Navas, Owen Gallagher, and xtine burrough (Eds.), //The Routledge handbook of remix studies and digital humanities// (pp. 109–124). Routledge. Dewey, John. (1934). //Art as experience//. First Perigee Printing. Friedman, Lester D. & Kavey, Allison B. (2016). //Monstrous progeny: A history of the Frankenstein narratives//. Rutgers University Press. Fyfe, Paul. (2018). Reading, making, and metacognition: Teaching digital humanities for transfer. //Digital Humanities Quarterly// 12(2). (link: "https://doi.org/10.63744/2hwpq24qx76m")[(goto-url: 'https://doi.org/10.63744/2hwpq24qx76m')] Greenberg, Ira. (2008). //Processing: Creative coding and computational art//. Apress. Hawk, Byron. (2018). //Resounding the rhetorical: Composition as a quasi-object//. University of Pittsburgh Press. Hunter, John, Faull, Katherine, & Jakacki, Diane. (2017). Reifying the maker as humanist. In Jentery Sayers (Ed.), //Making things and drawing boundaries: Experiments in the digital humanities// (pp. 130–138). University of Minnesota Press. Jagoda, Patrick. (2017). Critique and critical making. //PMLA// 132(2), pp. 356–363. Jasinski, James. (2001). //Sourcebook on rhetoric: Key concepts in contemporary rhetorical studies//. Sage Publications. (link: "https://doi.org/10.4135/9781452233222")[(goto-url: 'https://doi.org/10.4135/9781452233222')] TedX Talks. (2013, February 7). Paint your circuits with ink—A new revolution in electronics : Matt Johnson at TEDxGateway. //YouTube//. (link: "https://www.youtube.com/watch?v=R0xH0_qaqVw.")[(goto-url: 'https://www.youtube.com/watch?v=R0xH0_qaqVw.')] Johnson, Steven. (2022, April 15). A.I. is mastering language. Should we trust what it says? //New York Times Magazine//. (link: "https://www.nytimes.com/2022/04/15/magazine/ai-language.html")[(goto-url: 'https://www.nytimes.com/2022/04/15/magazine/ai-language.html')] Kashtan, Aaron Jacob. (2015). Materiality comics. //Digital Humanities Quarterly// 9(4). (link: "https://doi.org/10.63744/kn4rdcuhw8yb")[(goto-url: 'https://doi.org/10.63744/kn4rdcuhw8yb')] Mattress Factory. (2012, March 5). //Than Htay Maung - My Offering//. YouTube. (link: "https://www.youtube.com/watch?v=46d2xYku6m4")[(goto-url: 'https://www.youtube.com/watch?v=46d2xYku6m4')] O'Sullivan, Dan & Igoe, Tom. (2004). //Physical computing: Sensing and controlling the physical world with computers//. Course Technology Press. Peters, John Durham. (2015). //The marvelous clouds: Toward a philosophy of elemental media//. University of Chicago Press. Phillips, Bill. (2006). Frankenstein and Mary Shelley's 'wet ungenial summer.' //Atlantis// 28(2), pp. 59–68. Ratto, Matt. (2011). Critical making: Conceptual and material studies in technology and social life. //The Information Society//, //27//(4), pp. 252–260. (link: "https://doi.org/10.1080/01972243.2011.583819")[(goto-url: 'https://doi.org/10.1080/01972243.2011.583819')] Reynolds, Craig. (1999). Steering behaviors for autonomous characters. //Red3d.com//. (link: "www.red3d.com/cwr/steer/gdc99/")[(goto-url: 'www.red3d.com/cwr/steer/gdc99/')] Reynolds, Craig. (1987). Flocks, herds, and schools: A distributed behavioral model. //Computer Graphics, 21//(4), pp. 25–34. Reynolds, Craig. (1995). //Boids: background and update.// red3d. (link: "https://www.red3d.com/cwr/boids/")[(goto-url: 'https://www.red3d.com/cwr/boids/')] Rieder, David M. (2017). //Suasive iterations: Rhetoric, writing, and physical computing//. Parlor Press. Rizvi, Zahra, Chauhan, Rohan, Pue, A. Sean, & Zaidi, Nishat. (2022). Minimal computing for exploring Indian poetics. //Digital Humanities Quarterly//, 16(2). Russell, Legacy. (2020). //Glitch feminism: A manifesto//. Verso. Schumpeter, Joseph A. (1939). //Business cycles: A theoretical, historical, and statistical analysis of the capitalist process.// McGraw-Hill. Shelley, Mary. (1818). //Frankenstein; Or, the modern Prometheus//. Shiffman, Daniel. (2012). //The nature of code: Simulating natural systems with processing//. Free Software Foundation. Shipka, Jody. (2011). //Toward a composition made whole//. University of Pittsburgh Press. The Morgan Library & Museum. (2018, Sept 18). It's Alive! Frankenstein at 200 [Video]. //YouTube//. (link: "https://www.youtube.com/watch?v=0Q_Bq09rpEo&t=55s")[(goto-url: 'https://www.youtube.com/watch?v=0Q_Bq09rpEo&t=55s')] Tharp, Bruce M. & Tharp, Stephanie M. (2019). //Discursive design: Critical, speculative, and alternative things//. MIT Press. Thompson Klein, Julie. (2017). The boundary work of making in digital humanities. In Jentery Sayers (Ed.), //Making things and drawing boundaries: Experiments in the digital humanities//. University of Minnesota Press, pp. 21–31. Weiser, Mark. (1991). The computer for the 21st century. //Scientific American//, pp. 94–104. Woolf, Max. (2020). //textgenrnn// (Version 2.0.0). GitHub. (link: "https://github.com/minimaxir/textgenrnn")[(goto-url: 'https://github.com/minimaxir/textgenrnn')] [[Table of Contents]]

#For teachers and classroom implementation 0. ''Identify a text or piece of media'' you would like your students to engage with through a process of [[Creative Destruction]]. This could be //Frankenstein//—or it could be a work by Shakespeare, Jane Austen, Sylvia Plath, or beyond. 0. As an instructor, ''evaluate how you traditionally teach'' this text or piece of media. 0. ''Invite students to co-develop'' ideas, points of intervention, central questions, and methods with you. 0. ''Discuss and explore'' significant themes, and select a focused approach and topic together. 0. ''For low-tech implementation'', work as a class (or in small groups) to develop ideas and project proposals. Alternatively, students can create projects of their own design using ready materials. 0. ''For higher-tech implementation'', collaborate with your IT department, librarians, and/or YouTube videos to bring humanistic computing and ludic design to your classroom. 0. Importantly, ''discuss and reflect'' on the project's brainstorming, development, and creation processes as they are occurring. What is working well? What is confusing? How does this align with the messages of the text (or not)? 0. After the project construction, ''invite students to share and teach'' each other and you. [[Discussion questions]] [[How to make your own FrankenProject]] [[Table of Contents]]

#Discussion questions * How does the FrankenProject process complicate or clarify your understanding of the original work? * How do authors and/or characters engage in or enact [[Creative Destruction]]? * Where else do we see [[Creative Destruction]], and what form(s) does it take? * What does it mean to read and understand a text through a process of doing? In other words, what kind of reading or consumption is this? * How can literary themes expand beyond the text? * What other ways can we explore and study a text? [[For teachers and classroom implementation]] [[How to make your own FrankenProject]] [[Table of Contents]]

#The Pedagogy of Critical Making ##Calvin Olsen At its most basic, [[Critical Making]] is the process of thinking around and through concepts by creating a physical object—"learning by making," if you will. FrankenProject is very much a critical making project: Creating a physical interface that can be manipulated in real time by users helped frame our learning of Python and gave us new insight into many of the themes that run through Mary Shelley's //Frankenstein//. FrankenProject is an example of Matt Ratto's (2011) conception of critical making: It is a "goal-based material work" that "involves an iterative process of reconfiguration and conversation" that we and others can now reflect on (p. 253). And our experience with critical making offers a blueprint of how others might utilize critical making. One of critical making's (monstrous) traits is an ability to integrate itself into multiple contexts, often simultaneously. As Patrick Jagoda (2017) explained, "In academic discourse, [[Critical Making]] borrows from and intersects fields such as history of technoscience, human-computer interaction, digital humanities, and new media theory" (p. 357). Anne Balsamo (2011) added to that list by describing critical making's intersections with DIY culture, the maker movement, the creation of handicrafts, and modes of tinkering, modding, and hacking (p. 177). Critical making gets entangled with its contexts quickly and intricately, making it difficult to separate from those contexts. In not so many words, critical making is a lot of things sutured together—not unlike Mary Shelley's famous creature. Megen de Bruin-Molé (2021) wrote that Frankenstein's monster "is not incarnated perfectly anywhere, or at any time. Instead, it is everywhere, in many different forms. The monster is a process not a product; its power is collective and not singular" (p. 117). This is also an apt descriptor of critical making, and, like Frankenstein's monster, Ratto's formulation of critical making is difficult to pin down. Critical making is its own monster—a monster that, to use Meredith A. Bak's (2018) words, "reappears in different forms to more perfectly respond to and reflect particular constellations of social and cultural anxieties" (p. 223). One socio-cultural anxiety that [[Critical Making]] reflects and responds to is the status of the humanities in higher education. John Hunter, Katherine Faull, and Diane Jakacki (2017) showed that a "rhetoric of crisis" in the humanities goes all the way back to Erasmus—as does the idea of "making" in the humanities, as Erasmus argued that "ever since humanism existed as a self-conscious entity, its scholarly products have been fabricated" (pp. 261–268)—but they also proved that the number of humanities majors has declined since 2008 due to "a toxic combination of economic anxiety, spiraling costs, and media and parental suspicion of the humanities' value in the world" (p. 131). And they argue that critical making could be part of a response to the problem, stating that (align:"<==")+(box:"===XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX=======")[the idea that humanities students should be making real (rather than evanescent) products is a reasonable response to these pressures; it asserts the tangibility of what we do and addresses (in some cases, embraces) the demand to demonstrate 'relevance' in our historical moment. (p. 131)] Julie Thompson Klein's (2017) work shores up this argument, and she discussed the implications of the emergence of making within the humanities, writing that "it dismantles the dichotomies of doing versus thinking, production versus interpretation, and application versus theory," thereby "build[ing] authority for a new understanding of the complexity of making both within digital humanities and beyond, anchored by a growing body of literature and projects" (p. 27). Thompson Klein is quick to add that this paradigm shift does not equate to the jettisoning of previous practices (i.e., digitizing, archiving, etc.); instead, it resituates them as embodiments of creativity. A handful of scholars have begun using [[Critical Making]] as a pedagogical tool in their classes. For example, Paul Fyfe's (2018) work on digital humanities pedagogy in first-year experience courses showed that "critical making projects encourage the metacognitive mindset and process cycle which can lead to high road transfer" (p. 21). Fyfe's implementation and others like it are successful responses to the current iteration of the "crisis" in the humanities, and FrankenProject stands as another example of pedagogical critical making that other scholars can use to triangulate how the practice could be implemented in their own classrooms, and thereby, to quote Aaron Jacob Kashtan, find "an answer to the question of how digital humanities—or practice-based research, or critical making—could be made to work" for them (p. 2). If nothing else, we'd like FrankenProject to stand as proof that employing critical making in a classroom setting can be fruitful regardless of how "good" at making someone is. Zahra Rizvi, Rohan Chauhan, A. Sean Pue, and Nishat Zaidi's (2022) work on minimal computing and Indian poetics encourages practitioners to engage in jugaad, or frugal innovation, which "leans less towards engineering than bricolage" and allows critical makers to "inhabit a position closer to a bricoleur than to an engineer" (p. 11). This positioning of the critical maker as bricoleur maps wonderfully on to //Frankenstein// and Victor Frankenstein's dark foray into critical making. Although Victor Frankenstein's reaction to his achievement leaves much to be desired, his effective incorporation of multifarious pieces into a functional and fascinating whole is a process worthy of replication. [[Reclaiming Victor]] (keep reading) [[Processes]] (go back) [[Table of Contents]] (start over)

#Triangulating Transductions ##David Rieder In Shelley's //Frankenstein//, there are over a dozen descriptions of waves, both literal and figurative. Most of them are associated with natural disturbances, with wind and water. There are brawling waves, which break the silence of nature, and there are the waves on a troubled sea. There's a description of bare trees, their branches waving. There are rippling waves and waves that dash with fury. There is Victor's description of waves that roared and dashed at his feet while he walked along a stony beach. In fact, nearly half of all the descriptions of waves in the novel are characterized as dashing or restless. And the last sentence of her novel reads, "The monster was soon borne away by the waves and lost in darkness and distance." The prevalence of so many brawling, dashing, and restless waves in her story may have been due to the influence of the rain-soaked "Year without a Summer'' that Shelley and her friends experienced, which was caused by global climate changes linked to the eruption of Mount Tambora in Indonesia. Bill Phillips (2006) and others have argued that the horrible weather that kept Shelley and her friends indoors most of the summer of 1816 had influenced her deeply. Regardless of the origins, our interest in Shelley's waves is that there is a serendipitous parallel to them in the reliance on sine and cosine trigonometric functions in the Processing sketch for our project. Trig functions are the basis for the ways in which the curving stitches are drawn across the screen, which includes how the perpendicular lines are drawn to complete the simulation of the stitched pattern, the ways in which the animated lines of text move and transform, and how the COVID-19 cells float and rotate. They are also the basis for the flocking algorithm that is applied to the poem at the center of the project. In fact, the sensory data sent to the Processing sketch from the Arduino-based interface are a numerical representation of voltage changes caused by interactions with the hands comprising the interface, which are commonly visualized as a sinusoid. The reliance on sine and cosine functions in projects like ours is due to the ways in which they are used to simulate periodic waves along the y and x axes, respectively. Those waves offer programmers a "smooth" and continuous range of oscillating values that can be used to transduce a potentially infinite set of events and animations in any digital project, from curving lines and elliptical orbits to gradual changes in color, size, and alpha settings, to the behavior of particle systems, such as a simulated flock of birds. In //Processing: Creative Coding and Computational Art//, Ira Greenberg (2008) wrote: "trig functions are central to graphics programming" (p. 755). In an illustrated explanation of the way in which they are used, he reminded his readers that the trigonometric functions for sine and cosine let us "calculate the ratio of two sides of a right triangle" from an angle (p. 504). The sine function calculates the ratio of the opposite over the hypotenuse (soh), and the cosine function, which calculates the ratio of the adjacent over the hypotenuse (cah). The power of using sine and cosine functions as a transductive engine for generating periodic waves along the y and x axes is based on three relatively simple parameters of data: a constantly growing set of angular values, a fixed or changing set of values associated with the implicit wave's period, and a fixed or changing set of values associated with the wave's amplitude. As an example, using the sine function, if a developer writes a sketch that counts up continuously from 0 to 360 over a short period of time, and then continues to count up beyond 360 to 720 and then 1080 and then 1440, and so on, a continuously waxing and waning value will be produced. With each 360-degree turn, each full rotation around a computationally-defined circle, the diameter of which is defined by the period and amplitude applied to the equation, that value or magnitude grows and shrinks. Over and over again. And that dynamic value is applied as a transductive force to any of a variety of creative ends. [[Transduction]] (keep reading) [[Processes]] (go back) [[Table of Contents]] (start over)

#On Learning and the Mediated Glitch ##Kelsey Dufresne Victor Frankenstein is arguably the protagonist and antagonist of his own story: As he crafts life and brings his creation into being, he is simultaneously constructing a monster and his own demise—as well as the demise of everyone around him. This rudimentary overview of Frankenstein aligns with James Jasinski's (2001) claim that "individuals are both 'makers and made' or beings that create and get created" (p. 6). As Victor Frankenstein created his experiment, he was being fashioned as a god-like scientist that produced evil (again, a crude summary). Extending Jasinski's argument for its potential framing of this novel: We can simultaneously create and destroy—[[Creative Destruction]]. As Victor created and was created, he also destroyed and was destroyed. To bring his monster to life he destroyed various bodies (dead though they may have been) to bring them together in forming a new whole body. Upon bringing his creation to life, he immediately identifies it as a monster and abandons it, resulting in the monster then killing his loved ones and, ultimately, Victor dying in/from his efforts to undo his actions. Demonstrated here and throughout the entirety of the novel, Victor Frankenstein's actions teeter between creation and destruction so continuously that they became indistinguishable. Because of this, we were inspired to emulate Victor Frankenstein's methodologies in new forms—with text, machines, and visuals. As we have discussed, Mary Shelley's //Frankenstein// served as the channel through which we explored and practiced [[Critical Making]] production focusing on [[Creative Destruction]] and glitching. Here, we rely on glitch as our method, aesthetic, and framework for creation and exploration—looking to disrupt and distort as a means to explore, experiment, and engage with that which is familiar to efforts to find, uncover, or tease apart something new (or at least newly considered). We are inspired by Legacy Russell's conceptualization of glitch feminism in exploring the socio-technical constructs of gender and sexuality (2020), illustrating the datafication of the body and the tensions of self in our increasingly digital worlds: (align:"<==")+(box:"===XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX=======")[Yes, as abreu observes, we are indeed hyperlinks, signs and signifiers waiting to be clicked through, decoded, consumed. When we name bodies in an effort to make them useful, we end worlds, a process of codifying and delineating territory, limiting the capacity of the world around us and our agency within it. (pp. 76–77)] With Russell, we see how bodies are commodified, destroyed, created, and mediated through our individual and collective actions in the digital and the datafied world. Simultaneously, we look to Mark Weiser's (1991) claim on the immersiveness of ubiquitous computing: "The most profound technologies are those that disappear. They weave themselves into the fabric of everyday life until they are indistinguishable from it" (p. 94). We strongly divert from hidden, obscured tech—reveling in and revealing the glitch, including the wires, the hardware, the pointed materiality of technologies and computing to firmly situate a user in the collaged experience of these efforts. Through this project, we specifically explored the capacities of epoetry as rendered by natural language processing and GPT2 computer language learning to offer textual support to the multimodal experience centered on touch and visual manipulation. With natural language processing, we relied on the natural language toolkit (NLTK) to learn how to code in this fashion as a novice programmer. And we similarly relied on far more experienced coders with GPT2—which is a form of Artificial Intelligence (AI) where the computer is trained to produce its own textual artifacts. Yet, GPT2 carries an extremely problematic lineage as it has been found to produce racist, offensive, and problematic text as the materials in which it is trained with will result in works that carry and exasperate the biases and prejudices of its producers. With our project, we trained the computer with //Frankenstein// to explore what kinds of works and textual artifacts a computer would generate—what kind of textual glitches it would render, and what different FrankenPoems it would pull out. In doing so, we found that GPT2 is a technology and mediated representation of the text's core tensions, as well as that of our current concerns with technology: artificial intelligence, machine learning, flawed and oppressive algorithms, and the power structures that girdle our (increasingly digital) everyday lives. Thus, in relying on natural language processing and GPT2 to produce mediated pieces of text that effectively disrupted the originating work, //Frankenstein//, to a point of unrecognizability, we found that this process mirrored the [[Creative Destruction]] that Victor Frankenstein enacts on and through his monster. [[Creative Destruction and the Monster]] (keep reading) [[Processes]] (go back) [[Table of Contents]] (start over)

#Creative Destruction and the Monster Below is a code and output utilizing natural language toolkit—where the computer generated a variation of the text based upon the entirety of //Frankenstein//, read by the computer as a text file. In producing this piece in [[Collaboration]] with our computers, we were struck how this composition could be read and was expressive. Could this be a poem? Why not? #>>> franken.generate() (align:"<==")+(box:"===XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX=======")[me , and a presentiment of who it was , besides , I continued their single offspring . , or any other living thing . breathe the fresh air and bright sun or gentle breezes of spring ; all the hardships that I need. I hesitated a long , from every rougher wind and to perceive objects in their hands to direct my course towards the cottagers , their tumult is but as I first saw the lovely stranger , and one of his tale had kindled anew in me the greatest consolation that I knew it . this horror] We played around with the formatting: modifying, manipulating, and changing the form to present and perform, more visibly, as a poem. Effectively glitching this derivative. =====|===== me, and a presentiment of who it was besides, I continued their single offspring or any other living thing breathe the fresh air and bright sun or gentle breezes of spring ; all the hardships that I need. I hesitated a long , from every rougher wind and to perceive objects in their hands to direct my course towards the cottagers , their tumult is but as I first saw the lovely stranger and one of his tale had kindled anew in me the greatest consolation that I knew it . this horror [[this horror]] (keep reading) [[On Learning and the Mediated Glitch]] (go back) [[Table of Contents]] (start over)

#FrankenPoems Constructing poems and text pieces with GPT2 offered a vastly different experience compared to natural language processing. While we were not familiar with natural language processing, we were even less familiar with machine learning, which is precisely what GPT2 relies upon. Working with Max Woolf's (2020) "Interactive textgenrnn Demo w/ GPU," we were able to follow a program pattern of plugging in our text files, in this case Frankenstein, and modifying the settings of training to result in different and unique outputs. (text-colour:(hsl:120,1.0,0.35))[''Je rn a se ea ho lt oh e r t ee e h ed rh m e i a , sr d n g in t ln u e et ym aSe d p e ms o e er a tn v e eoo ne ee ec ew e i e p r e rm ede r nle mo fr d mr mn hd gt o e he a b ms e a d n eh e n res o t ehee I n n e il ai te g nr fe hl r s dt es rd s sa o td a et eos ae r n e sre u a eh kee ate t hte tct e e e gh b w he a n e sr bu t eu tc .s od ce nt h oo d oo ee eb e e rne nd n e ey ebi so e ir Im n d ih cm a r ,o hn cr e oe aie nh nh d ho ar soe e ua sr e n te e eir d s s ne a ye n neva oo r e o d er vh e e tt w ne n ca hs fso i n e p . tn n e m en fe ys ee t w h c efo ds E ne o te w y h ayo cm h yn io f a yn p t e e e owe dd eee re f l ie bt s os Iretthe co f t us i o are ye o ee s ua er d n eew r d e enth o e a e i ns ell f . eoe me r he rc s et s hn e Ae e ede r it or rt fd ho tn yn er i e se n e e r r ee d "i ee ge elh tn fn oe s eet sa le m e he vt tn r e oe ye us re l st di li e een as dc ts en o l e on ''] Above is an excerpt from a GPT2 FrankenPoem at an epoch training level of 1 (a lower level or extremely short interval of time for training). As you can see, the produced text is not readable—yet it is quite fascinating in form: the piece follows a paragraph-like structure, with spaces between what could be words, and even quotation marks demonstrating the computer's recognition of dialogue. For juxtaposition, below is an excerpt from a GPT2 FrankenPoem at an epoch training level of 200 (a much higher level or longer interval of time for training): (text-colour:(hsl:120,1.0,0.35))[''I was often tears in his glain occupations and promise with the work, it were the songs of their vines. I disamped to a heaver and more to be poor for ever. When they stranged them in that delighmtered more in your heart fly of your miserable devolaton, from meetimerels excipited as to the vain fieldise perhaps was conceived and was surprised to fled by an superiorn sounds with great cuition; but was take wrineced himself even vengeance. I should advise you from the cottage of hand De no down and hate your temper, passed in a stranger, and his eyes condemned lafting labours and power from witnesses, treaters are, felt still much at the crost morning, believes hut; bitter annimented and endeavouring to be supposed to its composuted." I loft that my powers existence, and the birth and knew that idea accompanied in contempt, your guilt was not even often that had low rest, with a fears, when I can returned to my faint distinguish, he was endeavoured to excellence by the trees or I could no loay remembered the rest; smiles or furture should be procast crept har thought; become sea to uttered a te returnity of the sailors lake, but it was but a plants our sound and solitary when I did notwy dicting the spirits of spirit, their enemy had not templessed by nor my fancy before friend; my trough and a much as his face and not?''] As the examples above illustrate, the longer the machine was allowed to train or learn, the more it created a readable output. The FrankenPoem generated with 200 epochs produced a far more readable, coherent, and clear work. While it is flawed with misspellings and points of severe confusion, it is certainly readable—much more closely aligned to work generated with natural language processing. These GPT2 poem pieces were utilized in FrankenProject as supplementary textual pieces, namely those that run in the background, whereas the poem generated by natural language processing served as the central text. What stands out to us as the most interesting result of these //Frankenstein//-derived textual pieces is how our FrankenPoems exist as glitched variations of Shelley's originating work—connecting them strongly to the monster as it was deemed such by his creator. Related to the glitch, if we were to further develop this project, we might explore the even more powerful and controversial aspects of GPT-3, the most recent iteration of OpenAI's Generative Pre-Trained Transformer. Steven Johnson (2022) critically explored the implications of GPT-3 in an essay titled "A.I. Is Mastering Language. Should We Trust What It Says?," which covers a lot of topics and territory. One topic that resonated with us was a section in which Gary Marcus is quoted: [Marcus] calls GPT-3 "an amazing version of pastiche generation, in a way that high school students who plagiarize change a couple words here or there but they're not really putting the ideas together. It doesn't really understand the underlying ideas." (Johnson) Johnson complicated Marcus' critique of the glitch-based origins of GPT-3 by focusing on how the intelligence attributed to GPT-3 evolves; so, the glitch at the heart of GPT-3 develops its own stitches. Like Victor's monster, it is able to learn to read and write on its own. We are especially fascinated to find that the credible texts GPT-3 brings to life echo both the theme of the glitch and the question of what a life really is, as well as to what extent we are responsible for the creations. And as a final note here, Elon Musk was one of the original founders of OpenAI in 2015. After his recent acquisition of Twitter/X, we wonder to what extent Musk, like Dr. Frankenstein, might be interested in plugging GPT-3 into the X-verse. How might "real" users of X unknowingly train artificially-intelligent users of X who've been birthed by a version of GPT-3 running in the social media application? [[The Visual Glitch]] (keep reading) [[this horror]] (go back) [[Table of Contents]] (start over)

#The Visual Glitch Similarly, we explored the capacities of glitch and [[Creative Destruction]] through visual collage(s). We reiterated this process of creative destruction with visual glitches (including distortions, disruptions, and embracing the messy and unexpected) and visual representations of key elements, scenes, and objects from the text. Operating in a reverse-annotation and searching for visual alignments, we then tethered various images and photographs together (yet again mirroring Victor Frankenstein's experience with and treatment of bodies). This was accomplished by finding images from Unsplash.com and then cropping, editing, changing, and deteriorating the files using the Windows Sketchbook program. Visually, the various scenes constructed, while not used in the final iteration of FrankenProject, provided a visual representation of the desired effect we hoped to achieve: a concrete sense of dissonance, an uncanny experience that situates a viewer between the tensions of the real and the unreal, and that which can be done to speak to the critical themes of the work at heart. Through this theme of glitching, we realized that the process and methodology of glitching—creating an intentional change in the intended form of both images and computer-generated poems—aligned perfectly with Victor Frankenstein's process of creating a living creature from different human beings/body parts. All the more, this process of employing [[Critical Making]] and creative coding to reimagine //Frankenstein// revealed numerous parallels between our pedagogical, technical, and thematic processes. [[Parallels between Shelley's work and our own]] (keep reading) [[FrankenPoems]] (go back) [[Table of Contents]] (start over)

#Parallels between Shelley's work and our own =|= ''Our work with FrankenProject'' * Our [[Critical Making]] method and product * Our collaborative work and remote community * Monster poems and visuals * Animated visuals with reliance on electricity * Tool's emphasis on user touch * User and GPT2 autonomy and agency =|= ''Mary Shelley's //Frankenstein//'' * Reading the Monster as a [[Critical Making]] product * Victor's Monster * The Monster's desire for community and family * Animated Monster brought to life with electricty * The autonomy and agency of the Monster * Monster created by Victor's hands, but also the death of many at the hands of the Monster and Victor |==| Our collective exploration into making a text unfamiliar (and making an unfamiliar text) mirrors the process in Shelley's novel wherein the human body, human capacity, and the self are all made and become unfamiliar. We now see how Victor Frankenstein engaged in [[Critical Making]] and production—centered on glitch creation and (mal)function—as his personal pedagogical processes at school. The defamiliarization of text, computer, art, place, literature, bodies, personhood, etc. illustrates how the methodologies of digital humanities and [[Critical Making]] can simultaneously produce both a project and a process when students learn technological activities and procedures first-hand. [[Processes]] (keep reading by choosing a different process) [[The Visual Glitch]] (go back) [[Table of Contents]] (start over)

#this horror In emphasizing the role of experience and experimentation here, as the method in which we all collectively and collaboratively learned as well as the intended mode through which other individuals will interact with and learn from our project, we look to John Dewey's framework of experience. Dewey (1934) defines art as "an object that belongs of itself in an enjoyed perceptual experience" (p. 92); therefore, situating experiences into art as expressive objects. Dewey identifies that a product of expression carries with it both an action and a result (p. 82), rendering an object expressive due to its ability to communicate to the viewer. Dewey further posits experience as that which is unified, complete with "continuous merging" (pp. 36–37)—but with unified cohesion, Dewey simultaneously identifies that art carries the capacity for expressiveness through abstraction and distortion: "Distortion of lines and departures from actual color may not only add to esthetic effect but result in increased expressiveness" (p. 92). Dewey argues: "Every work of art 'abstracts' in some degree from the particular traits of objects expressed. Otherwise, it would only, by means of exact imitation, create an illusion of the presence of the things themselves" (p. 94). Ergo, abstraction is inevitable, and through distortion the expressive object is better able to relay information through its communication without performing as the thing itself. We draw from Dewey in identifying the capacity of experience to further situate experiences as a force of abstraction and subsequent argumentation, which strongly aligns with Victor Frankenstein's own pedagogical experimentation that resulted in his monster—as well as that which resulted in these FrankenPoems. [[FrankenPoems]] (keep-reading) [[Creative Destruction and the Monster]] (go back) [[Table of Contents]] (start over)

#Glossary ##[[Collaboration]] ##[[Creative Destruction]] ##[[Critical Making]] ##[[Discursive Design]] ##[[Glitch Art]] ##[[GPT-2]] ##[[Mediated Reading (n.)]] ##[[Physical Computing]] ##[[Transductive Rhetoric(s)]] [[Table of Contents]]

#Creative Destruction Creative Destruction was originally a term coined by economist Joseph Schumpeter. Schumpeter (1939) meant for the term to be used in conjunction with modes of production, specifically the cycle in which new products and technologies disrupt, modify, and/or render obsolete entire sectors of production. We use the term more loosely in relation to the type of work that we (and Victor Frankenstein) embarked upon when we chose to learn via the creation of a physical object rather than working with "inanimate" technologies. It is our hope that others will continue the process by blowing up our version of creative destruction and using it to their own ends. [[Glossary]]

#Critical Making Critical Making is a term that was coined by Matt Ratto (2011) at the University of Toronto. At its most basic, critical making is the process of thinking around and through concepts by creating a physical object—"learning by making," if you will. FrankenProject is a partial critical making project: Creating a physical interface that can be manipulated in real time by users helped frame our learning of Python and gave us new insight into many of the themes that run through Mary Shelley's //Frankenstein//. [[Glossary]]

#Discursive Design Discursive Design is a process of engaging cognitively through the process of design—where designing becomes the opportunity and medium through which we can reflect on the world around us (Tharp & Tharp, 2019, p. 76). Here with FrankenProject, we engage in discursive design through our process of creating and designing to explore different ideas, themes, and tensions presented in Shelley's work—as well as the world. [[Glossary]]

#Glitch Art Glitch Art is visual art that has been created via the intentional manipulation of digital files. The aesthetic of glitch art often results in works that offer the appearance of malfunction, jumpiness, and instability. The type of glitch art employed by the FrankenProject included layering digital images on top of each other and writing code that caused the borders between images (and parts of images) to fluctuate when a user interacts with the touch interface. [[Glossary]]

#Physical Computing Physical Computing has been defined in various ways, but these various definitions can be distilled into a single broad one: physical computing is the creation of devices that can interact with their environment (i.e. heat sensors, induction loops, those freakin' cameras that send you speeding tickets, etc.). In not so many words, it is the process behind automation. In the case of FrankenProject, we created a haptic- or touch interface that can (1) read how much electricity is being pulled away from conductive paint when it is touched, (2) quantify that amount, and (3) react in specific ways to those amounts. In other words, the interface "knows" how much a user is touching it and "acts" accordingly. [[Glossary]]

#Transductive Rhetoric(s) Transduction, or the process by which energy is converted from one form to another, has been introduced to rhetoric studies by scholars, such as Casey Boyle, James J. Brown Jr, Steph Ceraso, Byron Hawk, and David Rieder. Each of these scholars values the rhetorical dimensions of transduction in different ways, but they share a general interest in the deeper, material processes that inform our rhetorical explorations of the digital. For example, for Boyle, Brown Jr., and Ceraso, (2018), transductive processes are the generative origin of digital culture, while Rieder (2017) values transduction as the basis for new forms of rhetorical invention and suasive action. Their approaches are different, but there is a shared interest in something before or beyond the screen. [[Glossary]]

#Making a Monster Early in the project when we were first connecting the Arduino to our computers, we decided to lean into the popular depiction of the monster in order to make FrankenProject immediately recognizable. Our prototype touch interface, which we called "Frankenstein Touch," included a green head with bolts in the neck. We purchased a Styrofoam mannequin head and spray painted it dark green. It was "life sized" when we started, and the Styrofoam kept its form so it remained recognizable as a head, but as the paint dried it caused the Styrofoam to shrivel slightly in places, thereby creating a textured head that appeared to have melted. Our first piece was born. Once the spray paint dried, a number of areas on the head were then painted again using Electric Paint, a water-based conductive paint made by Bare Conductive. The Electric Paint created a number of touch points on the head. Placement of the conductive paint splotches was performed sporadically, though we did take a note from Bare Conductive CEO and co-founder Matt Johnson's TEDx Talk (TedX 2013) and made a number of the splotches extra thick so that they could be used as capacitive sensors that could be hovered over rather than touched (3:23). The crown of the head received a coating thick enough to make it both a touch point and a place over which users could hover and still generate a reading, a splash across the right cheek resembled a scar, and both eyes were painted to up the ante on the "creepiness" factor. In fact, conductive paint is just viscous enough that applying it to the vertical plane of the eyes resulted in a tear or two running down the face. Despite being unplanned, this effect resulted in a welcome "emotional" aspect to the piece. After the conductive paint had dried, a hole was drilled on each side of the neck and two 1/2 in.-13 x 1-1/2 in. Zinc Plated Hex bolts were inserted. In addition to connecting the head to Karloff's monster and the immediate recognition of Frankenstein's monster, the bolts were also easily connected to the Arduino because they are made of metal. With the bolts in place, our monster head was complete. We used alligator clips to connect both bolts and every patch of conductive paint directly to the Arduino until every one of the twelve inputs on the MPR 121 sensor was occupied. Each connection registered on the screen with corresponding numbers even before they were touched—our monster had a mind of its own—and hovering over or touching each one caused the words on the screen to change. With the head completed and connected to the Python program via the Arduino, we were able to create a working FrankenProject prototype that could facilitate user-generated "readings" of //Frankenstein//. <img alt = "Head of FrankenProject with wires attached." src = "https://www.calvin-olsen.com/uploads/5/9/0/8/59087223/pxl-20201118-041431218-2-1_orig.jpg"> The head was a prototype of our touch interface, and the following process occured any time a user touched the Arduino: * the human body would pull energy from the interface via touch or hovering * the Arduino would register the amount of voltage lost to transduction and relay the information to the Python program * the Python program would perform a [[Mediated Reading (n.)]] of //Frankenstein// based on the new amount of voltage moving through the Arduino, and * a single word would appear at the bottom of the program The touch interface was sensitive enough that the word appearing at the bottom of the screen almost always appeared unstable—glitchy, if you will—bouncing back and forth between a handful of words in rapid succession. The unstable readings could be manipulated to a degree if the user modified the number of locations they were touching simultaneously and/or changed the amount of pressure placed on a particular touch point; but, regardless of what modifications users chose, it was nearly impossible to keep a particular word or series of words on the screen for more than a moment. This made for some interesting connections to //Frankenstein//. Among others 1. By touching the interface, users became literally connected to the //Frankenstein// text, especially to Victor's inability to control the monster once it came to life and his journey from creator to spectator as the monster reacted to his decisions. 2. The physical appearance (see also: creepiness) of the head and the requirement that users touch it mirrors for some new users the trepidation Victor felt upon seeing the creation for the first time (its yellow eyes being particularly unnerving). 3. The random connections between words that appeared on the screen would at times mimic or call forth themes and ideas existent in the text. For example, in one of our earliest tests the words "my" and "father" flickered on the screen repeatedly after we were no longer touching the head and did not let up until another location was touched. [[Getting Out of Hand]] (keep reading) [[The Pedagogy of Critical Making]] (go back) [[Table of Contents]] (start over)

#Getting Out of Hand With the head-based prototype finally working as planned, we brainstormed ideas for other body parts that might be appropriate to include in an expanded touch interface. The Styrofoam mannequin head had worked so well that we sought out other mannequin pieces. We started by purchasing a plastic mannequin torso. It was a male torso that ran from the base of the neck to the mid-thigh with no head and no arms. Because it was made of heavy-duty plastic and had a metal core, it had a much stronger verisimilitude to the human body in both shape and weight, to the point that it felt strange carrying it around in public. Though it would have been possible to modify it with spray- and conductive paint to appear more like the head we had created, we ultimately decided that the torso's appearance would distract users and make them less likely to explore various touch points. With user experience being high on our priority list, we scrapped the torso and opted instead to focus on appendages as vehicles for creating a more inviting interface. Appendages were conceptually interesting to us because using only appendages would render the body a negative space, and negative space connects the physical elements of the project to the poetry that it was generating. After a few short and unsuccessful forays into arms and legs, we came across the work of Burmese artist Than Htay Maung, whose installation at the Mattress Factory Art Museum in Pittsburgh in 2012, "My Offering," stands as a singular demonstration of the rhetorical power of hands. The installation is made up of numerous plaster hands (individually cast by the artists using his own hands as model) protruding from walls. Each hand holds a piece of bread, which Maung explains comes from the relief work he did for victims of the cyclone that devastated the Burma delta in 2008 (the event that pushed him and his family to move). Describing the installation, he said, "Whenever we would come to an affected area, there would be all of these hands outstretched asking for food, for safety, for kindness and for others. I cannot escape those scenes—the view of those hungry hands" (Mattress Factory, 2012, 0:53). This description reminded us of the monster's plight in Frankenstein when it is living in a hovel and trying to survive in dire circumstances outside of its control. Additionally, the terms "hand" and "hands" appear far more often in Frankenstein than almost any other body part (with the exception of "eyes"). Deciding on hands also worked in conjunction with some of the parameters of our project. Even a number of hands would take up less space than a torso, and hands would be more easily created, individualized, and transported than other body parts. And since we were planning to make the interface appear as a surface in Victor's lab might, a series of hands allowed us to match Victor's process: His use of a hodgepodge conglomeration of parts to build a body of disparate parts is mirrored in our choice to have the hands made out of a material with a number of characteristics extremely different from those of Styrofoam. Hoping to round out this journey through [[Critical Making]] (our first) by learning another new process, we looked to the Makerspace at NCSU Libraries and chose to take advantage of their 3D printing services. We were unable to frequent the Makerspace at NC State's Hill Library in person because of social distancing protocols, so we opted for finding a 3D model of a hand online. We found a model intended for displaying jewelry at Thingiverse.com made by user vicentegasco that fit the size requirements of the Makerspace's 3D printers. We ordered five copies of the model at 50% infill so that they would be able to withstand some force, and we requested that at least one copy be a shade of green in order to connect it (however slightly) to the Styrofoam head. A few days later we collected the five hands—one dark green, two red, one yellow, and one semi-transparent—up at the library help desk. When these five hands were ready to be part of the touch interface (process described below), we utilized FreeCAD software and online tutorials to help us mirror the image in order to produce a right hand, thus allowing us to include five pairs of hands rather than ten left hands. The right hands were mirror images of the left, but there was much less variation in color (one bright green hand and four gray hands) and we decreased the infill from 50% to 30%. As mentioned, we had originally chosen 50% infill for the original batch because we wanted to make sure that the hands were sturdy enough to stand on their own, especially after being drilled through and covered in heavy conductive paint. After receiving the first batch back from the Makerspace and preparing them for use, we knew that vicentegasco's 3D model was sophisticated enough to hold up to repetitive use even without a high infill. Ergo, we dropped to 30% (the lowest possible option for the 3D printers to which we had access) to reduce our environmental footprint and make our lives easier with the drilling. It was interesting to see how the 3D printing process was affected by the change in infill percentages in the 3D printing. We were under the assumption that less plastic meant faster printing; however, the printing process took longer after we dropped the infill percentage. The five left hands from the first batch were ready in three days and had almost no defects whatsoever, while the five right hands took well over a week to finish printing and each came with a number of structural and cosmetic flaws (missing lines, rough edges, sharp fingernail fragments, etc.). The main pitfall of not being able to design and print the hands during a live session at the Makerspace was that we were unable to procure the results of failed jobs. The model was complex enough that each hand took roughly nine hours to print. Since nothing seemed to go awry with the first round of printing, we were oblivious to the fact that the 3D printer would sometimes cancel a printing part way through, leaving an incomplete and/or mangled hand that would have added interesting dimensions to our project's aesthetic. <iframe src="https://docs.google.com/presentation/d/e/2PACX-1vTvooLnKFwA2n6OGAbVYvX0r3yCVvfu515aU_ML8bmT5cF23s6Vf1ifFltYusNc1RP8MFpcmk57rikU/embed?start=true&loop=true&delayms=10000" frameborder="0" width="960" height="569" allowfullscreen="true" mozallowfullscreen="true" webkitallowfullscreen="true"></iframe> [[From Python to Processing]] (keep reading) [[Making a Monster]] (go back) [[Table of Contents]] (start over)

#From Python to Processing Connecting the hands to the touch interface was a much more complicated undertaking than connecting the head to the Arduino. It was much more time consuming to paint and drill hard plastic than it was Styrofoam (though, for better or for worse, no "melting" occurred). And the scope of our project had expanded significantly by the time we were creating the hands and modifying them to connect to the rest of the [[Physical Computing]] apparatus: The touch points on the hands were more complicated to wire for continuous conductivity, we expanded the number of input sensors available for generating readings (which were also expanded from only text to include, images, movement, and sound), and we had switched from a Python-based program to a Processing-based program. The reason for the switch from Python to Processing was ultimately due to limitations associated with complex graphical and textual animations and visualizations. Python has capacities for both visualizing data and for relatively simple animated games. It can generate complex visualizations from data, and the visualizations can be animated, albeit in limited ways. Python also has the capacity to work with the sensory data from an Arduino microcontroller; nevertheless, Python proved to be insufficient for our goals. We experimented with PyGame, which is a set of Python modules designed for creating computer games. We were able to animate graphical assets in a simple gaming environment from interactions with an Arduino touch sensor, but the complexity of the animated, multimodal animations that we hoped to create proved too complex for PyGame. Processing, which was designed to work with Arduino, allowed us to accomplish everything that we'd hoped to implement. We started by giving the 3D printed hands a coat of conductive paint and doing our best to make each one unique in some way. For example, one or two received a full coating of paint, one had only the fingertips painted, one had the palm and back of the hand painted with the fingers left entirely alone, etc. Each hand was printed with a base so as to allow it to remain upright, and any hand with conductive paint extending to its wrist had at least one hole drilled through it after the first coating of paint. A dollop of paint large enough to make it through the wrist was applied to this hole using a Q-tip, and the hand was then tipped over so the paint coming through the hole could be spread across the inside of the hollow wrist for easier application of adhesive. After the new paint dried, a foot or two of 300 volt wire was stripped at each end so one end could be inserted into the breadboard and the other end could be secured to the paint inside the wrist using copper tape and, when necessary, conductive wire glue (manufactured by Anders Products). The few hands that did not have paint at the wrist were more complicated, but they also allowed for the creation of multiple touch points. The index, middle, and ring fingers of each hand were hollow enough near the base that they could be drilled through without compromising the structure of the hand. In these cases, the wire was run through the hole from the outside, allowing for the end not going into the breadboard to be bent so that it caught on the base of the finger. Instead of taping these ends, a second (and particularly thick) coating of paint was used to cover the wire—the conductive paint was viscous enough to stay in place while drying and render the few millimeters of wire essentially invisible near the space between fingers. With the head and all of the hands wired to the breadboard (and, by extension, the computer), all of the FrankenProject touch interface pieces are ready for assemblage. Materials used for the FrankenProject head-based prototype * Mannequin head (Krylon Fusion ALL-in-One Paint+Primer, Matte Spanish Moss color) * Conductive Paint (Electric Pain made by Bare Conductive) * Alligator Clips (ELEGOO Double-End 50cm Jumper Wire with Alligator Clip Text Leads) * Hex Bolts (2) (1/2 in.-13 x 1-1/2 in. Zinc Plated Hex Bolts) * Arduino (Arduino Mega 2800) * Wire Glue (Anders Products Conductive Wire Glue) Materials used for the complete touch interface for FrankenProject * 3D Printer (Provided by NC State University Libraries) * Drill bits (30-50% infill) * Drill (Ryobi HJP003 12V Drill Driver) * 3D Printer Filament ("Do it" brand titanium drill bits, sizes: 1/16, 5/64, 7/64) * Copper Wire (22 gauge AWD) * Wire Glue (Anders Products Wire Glue) * Copper Tape (Generic, any brand/size works) * Super Glue (Generic, any brand/size works) [[Processes]] (keep reading by choosing a different process) [[Getting Out of Hand]] (go back) [[Table of Contents]] (start over)

#Transduction The concept of transduction is essential to an appreciation of the role that the computationally driven waves represent in FrankenProject and many other digital projects in which trig functions are used. To understand what the concept means, we can begin by appreciating that it is an interdisciplinary concept, and one that has appeared recently in our field. In //Physical Computing: Sensing and Controlling the World with Computers//, Dan O'Sullivan and Tom Igoe offered a simple description that begins with the following: "One of the main principles behind [[Physical Computing]] is transduction, or the conversion of one form of energy into another" (2004, p. xx). Physical computing is a generic term for the smart, interactive, post-desktop paradigm of popular computing in which we are increasingly engaged today. It is the paradigm that has led to studies of hybrid and mixed realities as well as to concepts such as eversion. O'Sullivan and Igoe went on to explain that the process of transduction is "about creating a conversation between the physical world and the virtual world of the computer" (2004, p. xx). Finally, writing about the goal of this "conversational" process, they explained, (align:"<==")+(box:"===XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX=======")[The process of transduction, or the conversion of one form of energy into another, is what enables this flow. Your job is to find, and learn to use, transducers to convert between physical energy appropriate for your project and the electrical energy used by the computer. (pp. xix–xx)] In a project such as ours, which is an example of a [[Physical Computing]] project, an example of the kind of transducer to which O'Sullivan and Igoe are alluding is the MPR 121 capacitive touch sensor. Those sensors, as with all sensors, are designed to transduce the physical energy for which they are designed in ways that can then be used by a computer system. The MPR 121 sensors detect changes in the sine wave of voltage associated with each of its inputs. Since the human body is electrically conductive, when a user touches our interface, they draw down the voltage running through one or more inputs. The voltage change is a transduced conversion of the energy associated with tactile touch. The sensors are the basis for the conversation on which our [[Physical Computing]] project is based. In their article, "The Digital: Rhetoric Behind and Beyond the Screen,", Casey Boyle, James J. Brown Jr., and Steph Ceraso (2018), offered a definition of transduction that echoes the one offered by O'Sullivan and Igoe, but a key difference arises in their application of it in our field. In the final section of their essay, "Tomorrow's Rhetoric, Transduced," they wrote, (align:"<==")+(box:"===XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX=======")[Transduction refers to how a signal moves across disparate registers of relations: neural firings move to fingers to perform keystrokes that then transform into electrical charges that then become digital bits and are delivered to a screen by software or saved to a hard drive that becomes transcoded again whenever someone opens a file. (p. 257)] But, for Boyle et al., transduction is a far more abstract and universal attribute of "the digital" today: They explain, "neither deductive or inductive, we define the movement of the digital as transductive" (p. 257). What they mean is that "the digital" is no longer something separate and distinct from the real or physical world; so, a transductive process isn't just about creating a conversation among transducers to mix the two realities. They argue that transduction is an abstract, universal attribute of the digital. Transduction is an essential part of what John Durham Peters would call a condition of our existence. It is now infrastructural, a "form of life" (Peters, 2015, p. 14; Boyle et al., 2018, p. 256). Boyle et al. conclude, "the digital portends to be a momentary specialization that falls away and becomes eventually known as the conditions through which rhetorical studies finds itself endlessly transducing" (p. 258). Transduction is an attribute of the digital, and as the lines between the virtual realm of the computer and the "real" world increasingly blur—mixing, hybridizing, and everting—our ability or interest in "the digital" will fall away. In an age of [[Physical Computing]], of smart, interactive objects and environments, rhetors will turn their attention to a search for the available means of transduction. [[Co/Sine Waves and Transduction]] (keep reading) [[Triangulating Transductions]] (go back) [[Table of Contents]] (start over)

#Co/Sine Waves and Transduction When the implicit circles, which are drawn over and over again by a block of code, are passed directly or indirectly to trig functions in our Processing sketch, transducible waves of energy are generated, which can be applied to a potentially endless array of animated sequences and designs. The figure below is a screen capture in which several transduced effects are visible: the COVID-19 cells, the stitches, and what we've been calling the text noodles. Each of those animations are based on direct or indirect applications of trig-generated waves. <iframe src="https://docs.google.com/presentation/d/e/2PACX-1vTdhJuOcMeAlMK3nrRqPsLOo7zrP474Xswm2lXF16CRYmhUt1cR9r9wudiNP6TcTmSmQJxF6FQCtK9F/embed?start=true&loop=true&delayms=60000" frameborder="0" width="960" height="569" allowfullscreen="true" mozallowfullscreen="true" webkitallowfullscreen="true"></iframe> //Placement of stitches, text, and COVID-19 cells based on trigonometric functions.// Each of the COVID-19 cells floating around on the screen is based on the same .png. But the elliptical orbit that each one follows, each cell's slow spin around its center point, their growing and shrinking size, and the waxing and waning brightness that each one exhibits, is the application of a transduced sine wave of oscillating value. And slight differences in ways in which the transduced sine wave is applied to each of the above-mentioned characteristics of the cells gives them a more realistic appearance in the project. The perpendicular lines comprising the stitches are also reliant on transduced trig values, albeit the approach is different. In a function that we wrote titled stitchMaker(), four functions in the Processing programming language are used to place and properly rotate a short line comprising the simulated stitches in our project. The functions are curvePoint(), curveTangent(), atan2(), and rotate(). To understand how curvePoint() is used, it's important to understand that the long line comprising the stitches, which are drawn across the screen, such as the wavering horizontal line in the figure below, are drawn computationally from a list of two-dimensional points. What curvePoint() does is to determine a point among a subset of the list of two-dimensional points where a perpendicular line can be drawn. That determination is based on implicit trigonometric functions. Once curvePoint() determines where the perpendicular stitches can be drawn, curveTangent() and atan2() are used to determine the perpendicular angle. Since the long lines comprising the stitches waver, changing their angle, the angle of the perpendicular stitch line wavers, too. Finally, the rotate() function applies the calculated angle to the shorter lines, which are then drawn to the screen. <iframe src="https://docs.google.com/presentation/d/e/2PACX-1vSZLeVv6sLWzpUd2aQXpMd7bLQm4fcKu4JqZ6TuodNuAeM2gmhDOuZUTDRTDCFW0Bq4f3xUhoDiY4gA/embed?start=false&loop=false&delayms=3000" frameborder="0" width="960" height="569" allowfullscreen="true" mozallowfullscreen="true" webkitallowfullscreen="true"></iframe> The rotation of the letters in the text noodles is based on the same set of functions listed above. Each letter is placed along an implicit line on the screen, and as the line twists and turns, the four above-mentioned functions are applied to each letter so that they are placed at the correct rotation. Since the text noodles are animated, as they move along the implicit lines defining their paths, each letter takes on the rotational values of the letter preceding it in the line. The text noodles are another example of the waves upon waves of transduced energy that has ultimately made its way to each of the visual and kinetic forms in our project. The sounds in FrankenProject are based on seven .mp3 and .wav files which were extracted from YouTube videos and then cut down and optimized in Audacity for Processing's sound library. All of the sound files are held and referenced in an array, and two or more of the seven files play simultaneously on a loop while FrankenProject is running. Two arrays of boolean values, which parallel the length and order of the sound array, are the basis for managing the process of playing/pausing each of the sounds. Touch-based user interactions with our project change the amplitude or volume for each active sound, and a timer—which counts the milliseconds of play since FrankenProject started, and which is used to change the layered background images in our project—is also used to progressively swap out one sound for another. A final example of the application of wave-like, trigonometric functions in our project is associated with the GPT-generated poem. In the figure above, the poem is displayed in our project as a conventional two-dimensional object. The reader may not realize that each letter in each word is an object in a particle system, and that each letter-as-particle is waiting to be activated, at which point they each participate in an algorithmic design of behavior that is meant to simulate flocking. In the figure below, the effects of the algorithm on the position of the letters and the relationships among the letters comprising the poem is apparent. Not only have the letters in the poem broken free of their conventional grid-like formation, floating away, but they have become to simulate the characteristics of flocking birds. <iframe width="560" height="315" src=" https://www.youtube.com/embed/8lCbgiYsTRA?si=s7xUAgXKvq29T3VV; YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe> The algorithm on which the simulated flocking behavior was adapted for our project is Daniel Shiffman's Boid class, which he wrote in Processing. In //The Nature of Code: Simulating Natural Systems with Processing// (2012), Shiffman explains that his class is based on the work of Craig Reynolds, who, in 1987, published a paper titled "Flocks, Herds, and Schools: A Distributed Behavioral Model." In that paper, Reynolds explains how he and his team addressed the challenge of simulating the decentralized dynamics of animals, such as birds, as they flock and move together. Earlier attempts in computer graphics to portray a flock of birds scripted the paths of each individual bird. His solution was to create a particle system in which each particle or bird acts like a simple, virtual computer, making decisions for itself, based on computational behaviors programmed into it. In a later essay in which he follows up on the argument in his 1987 paper, Reynolds's (1995) terms for flocking behavior were 1) separation, 2) alignment, and 3) cohesion. Our repurposing of Reynolds's flocking algorithm was another way in which we were able to extend stylistically the //Frankenstein// trope of the dis/unified body to the text at the center of our project. The behaviors for each individual letter in the text could be changed by user interaction with the hands comprising our physical interface. Each individual letter, like the body parts comprising Frankenstein, have past histories and trajectories of their own, and our use of the flocking algorithm was a way in which we could embed within each lettered "body part" in our text its own behavior, its own desire. To build a simulated flock, we start with a boid model that supports geometrical flight. We add behaviors that correspond to the opposing forces of collision avoidance and the urge to join the flock. The behaviors that lead to simulating flocking are: * Collision Avoidance: avoid collisions with nearby flockmates * Velocity Matching: attempt to match velocity with nearby flockmates * Flock Centering: attempt to stay close to nearby flockmates Related to the theme of waves and trigonometry, these geometrical behaviors are computed as what Reynolds (1999) calls a "desired velocity": a vector that reconciles the individual, curvilinear flight path of a boid with the paths of its nearby flockmates. In his 1995 follow-up essay, "Boids: Background and Update," Reynolds wrote, "Each boid has direct access to the whole scene's geometric description, but flocking requires that it reacts only to flockmates within a certain small neighborhood around itself" (paragraph 2). That reconciliation is based almost entirely on trigonometric functions that are able to reconcile different flight paths by determining the difference in flight paths by connecting them as points in a right triangle and then using trig functions to identify the angle that defines their difference. The individual boid's behavior is programmed to shrink the angle that defines the difference between the individual's path with that of nearby flockmates in order to join them. It is akin to lessening the period and amplitude of the implicit wave that separates one boid from another. So, the waves that dash and brawl and break throughout Shelley's novel are represented in our project, too, albeit at the more abstract level of code and computational methods. [[Processes]] (keep reading by choosing a different process) [[Transduction]] (go back) [[Table of Contents]] (start over)

#Collaboration At its most dictionary-esque, collaboration is working together. We like to think of it as making something together. We also like that collaboration can refer to a thing that has been created by multiple people. And since FrankenProject is a thing that we three made together, it is a collaboration collaboration. And since the pieces it's made of—materials, code, ideas, etc.—are themselves conglomerates of ofttimes disparate pieces, FrankenProject is a collaboration collaboration collabo—you get it. [[Glossary]]

#Mediated Reading (n.) Mediated reading is a term we use to describe the actions FrankenProject takes once it is provided with input (touched). What FrankenProject does can be considered a reading because words are moving from the text into the mind of a human participant. And this reading—this action—is mediated because its existence depends entirely on the physical materials used to spark it. In not so many words, FrankenProject performs mediated readings because it chooses pieces of the text for you rather than you choosing what pieces to read. And though users can manipulate what appears on screen by changing how and where they touch the interface, they are unable to proceed without interacting with the physical media. The authors realize this glossary entry is a can of worms, but we also hope this has helped a bit. [[Glossary]]

#(Re)claiming Victor The first question we asked ourselves upon embarking on this project was "Is it possible to create an experience that connects users to both Victor and the monster?" This question was interesting to us because the cultural consciousness of //Frankenstein// is so heavily influenced by the conflation of Victor Frankenstein and his creation. Many readers in the United States meet //Frankenstein// in high school English courses, but popular depictions of the monster are largely separated from Shelley's novel. The monster has appeared in films since 1910, the year Edison Studios produced the silent film //Frankenstein// in which the monster was portrayed by Charles Ogle, and the depiction of the monster as a flat-topped being with a "fivehead" and bolts in its neck was cemented in the popular imagination once Boris Karloff starred as the monster in James Whale's films in the 1930s. Karloff's monster has become the quintessential image for Victor Frankenstein's monster, as any Google image search or screening of the 1999 cartoon film //Alvin and the Chipmunks Meet Frankenstein// will attest. Because of this pop culture history, casual use of the term "Frankenstein" almost always refers to the monster, so extracting Victor was of particular importance to us. We found that the process of creating FrankenProject connected us to Victor in three ways. First, we sought out and experimented with different "body parts"—individual pieces that we believed would be functional as a conglomerate despite their not being uniform. Second, this was very much our first attempt at building anything of this type, and much of our work was performed via trial and error. We were well beyond our depth at times, but our vision for what the final result of our work would be expanded as we gained and honed new skills. And third, the scope of FrankenProject very quickly expanded well beyond our original plans. As the interface and computer program increased in sophistication, we were called upon to make adjustments in an attempt to corral the thing we had made. And this attempt was driven by our own intellectual development as much (if not more than) by the opportunity to make a thing that works. In this section, we will describe the making process that took us from a working prototype of a broad desktop-based FrankenProject to a complex immersive experience, along with a list of the specific materials we used. [[Making a Monster]] (keep reading) [[The Pedagogy of Critical Making]] (go back) [[Table of Contents]] (start over)

#GPT-2 GPT-2 (Generative Pretrained Transformer 2) is a language model developed by OpenAI that generates human-like text by predicting the next word in a sequence. It was trained on a large dataset of internet text and can produce coherent and contextually relevant responses to prompts without task-specific tuning. In the FrankenProject, GPT-2 was used to generate text fragments that were unpredictable and varied, contributing to the project's themes of fragmentation and recombination by offering outputs that echoed both sense and disorientation. [[Glossary]]