This paper seeks to investigate ChatGPT from a narratological perspective, specifically asking the question: Is ChatGPT a narrator? There is a growing body of qualitative research that suggests Large Language Model (LLM) technologies have the capacity for dishonesty and inaccuracy. This paper will attempt to contribute to this avenue of investigation but from a humanities perspective. I will do so by employing a methodology of narratological analysis to asses whether ChatGPT is a narrator and, if so, to what degree it is reliable. A significant proportion of the data used to train ChatGPT’s LLM was drawn from the history of literature that was scraped from the internet. Therefore, there is an urgent need for an exploration of ChatGPT’s textual outputs via narratology. This paper intends to make contributions to critical AI studies via the field of narratology by working toward a broader narratological theorisation of ChatGPT.

During World War I, Australian soldiers trained in either Egypt or England, before serving in Gallipoli, the Western Front, or the Middle East. Despite these common locations, and path to war, each soldier had a personalised journey. This journey depended on a variety of factors, including being assigned to different units or roles, leave, injury, being taken prisoner, and death. These individual journeys are highlighted in their war diaries. The aim of this work is to automatically map each soldier’s journey through the war using named entity recognition (NER). The diaries used in this analysis come from the State Library of New South Wales. The initial step in this work is extracting locations using NER and cleaning them. However, this is difficult due to a variety of factors, including misspellings, shortened location names (e.g. “Aus” instead of “Australia”), generic locations (such as “town”), and the extraction of non-locations by the NER. This paper will discuss our work to date in using NER and dealing with the aforementioned issues. We will also present initial ideas regarding geocoding these locations. Geocoding allows us to obtain the coordinates of a location and easily map it. However, this can be difficult as many locations share names (e.g. there is a Liverpool in both Australia and the UK). Furthermore, diarists may have mentioned locations they were not at. Here, we hope that the temporal information of the diaries can aid us in correctly identifying each location.

In the second note that appends her translation of L. F. Menabrea’s "Sketch of an Analytical Engine" (1842), Ada Lovelace describes how the “results” of Babbage’s computing machine would present themselves on the “perpendicular face of the engine [. . .] opposite to which face a spectator may place himself” (28).* This paper considers Lovelace’s speculative detailing of the encounter between the Analytical Engine and the viewing subject and thinks through how it might model a form of close reading that is familiar to the field of literary studies. In contrast to the type of reading that the digital humanities make possible (e.g., big, vast, distant), I suggest Lovelace’s “Notes” also offer a close-up view of a mechanical computer that prefigures our relation to the electronic interfaces that mediate reading in our contemporary moment. Drawing on Johnathan Kramnick’s (2021; 2023) analysis of the nexus between close reading and craft, I explore the resonances between Lovelace’s nineteenth century reading of the Analytical Engine—in particular her explanation of its technical connection to the Jacquard-loom—and the practice of close reading today. Further, the paper hopes to offer a brief comparative reading between Lovelace’s encounter with the Analytical Engine and some other, more contemporary, critical accounts of reading machines [i.e., Jacques Derrida’s Paper Machines (2005) and Andrew Piper’s Book Was There (2012)] that might help clarify a reading of Lovelace’s “Notes” that recognises her as an early reader of the digital page.

*Babbage, Charles. “Article xxix., from Scientific Memoirs, vol. iii., p. 666.” "Babbage’s Calculating Engines: Being a Collection of Papers Relating to Them; Their History and Construction," edited by Henry P. Babbage. Cambridge University Press, 2010, pp. 4–50.

Speculative fiction (SF) is rife with fraught representations of digital technologies. From its early history, it has posited narratives of techno-pessimistic cautionary tales of ethical quandaries and the sheer hubris of innovation, or valorised techno-optimist futures in which novel technologies absolve us of responsibility to land and to each other. In this presentation, I raise a throughline in my thesis-in-progress about diasporic and Indigenous futurisms, regarding the speculative ethics of digital archives in relation to cultural maintenance and wellbeing. I examine the digital-embodied dynamics in two recent works of SF, Goori and Lebanese writer Mykaela Saunders’ <em>Always Will Be</em> (2022), and Grace Chan’s <em>Every Version of You</em> (2023). I explore how these texts stage encounters at the interface of the embodied and the digital, raising questions about how digital data practices recover, facilitate, or erase cultural knowledges. I read Saunders’ portrayals of community-centred archival sovereignty, learning from her recontextualisation of archival power, before turning to Chan’s portrayal of diasporic archives falling through the cracks, where translations between the qualitative realm of culture and the quantitative world of digital data is fraught, and culture cannot so seamlessly traverse the boundary between “meatspace” and the digital virtual. This presentation explores how these speculative ethics of the archive reimagines our relationship with digital data and asks: how do these technologies facilitate or foreclose cultural knowledges?