I’m in Australia! Hooray! My mum and dad and sister and I moved from Perth to Bergen when I was a kid, and ever since I’ve loved two homes: Australia and Norway. Walking off the plane last night and breathing in the scent of eucalyptus in the warm night air my whole body relaxed. I’m home.

I’m giving talks here in Perth at Curtin on 24 November, in Sydney at UNSW on 26 November and I’m keynoting the Australasian Digital Humanities conference in Canberra on 4 December with a talk titled Archipelagos or Empires? Narrative Colonialism in Generative AI. I’m also meeting Crystal Abidin‘s team at IERlab at Curtin University for one of their Tea Sessions, and a few other people.

Please get in touch if you’re in any of these places and interested in the AI STORIES project! My email is jill.walker.rettberg@uib.no.

Here’s more about the talks:

Monday 24 Nov, Perth: The Politics of Researching Synthetic Media: What Is Good Internet Research If the Internet Is AI-Generated?

Public talk at CCAT at Curtin University. Here’s the announcement!

– Date: Monday, 24 November 2025
– Time: 12pm – 1:30pm (includes time for questions and discussion)
– Location: Room 204.233 or online
– Parking: Nearest parking is Pi1 or Pi2
– Light refreshments provided; feel free to bring your lunch
– RSVP: Please RSVP for in-person attendance to MCASIadmin@curtin.edu.au – there is also a streaming link that you can get from the same email.

Abstract: More and more of the internet is AI-generated. If you Google anything, you not only get an AI-generated summary at the top of the results—many of the top hits are likely AI-generated webpages. On social media, humans enhance their photos and comments with AI suggestions and edits, or generate all their content. Some platforms, like Butterfly.ai, are entirely run by bots, and chatbots like Replika or ChatGPT are happy to be your friend. Some worry that, soon, language models will be trained on AI-generated text—and the internet will eat itself (dead internet theory).

What does it mean to research this algorithmically generated cultural content? Is AI just another tool, another medium? Should we think of it as a human-AI collaboration? Or are humans increasingly irrelevant? Do we need to know? And is it OK for human researchers to use AI to research online culture?

After nearly a decade of researching AI, Jill Walker Rettberg has become increasingly frustrated. In this talk, she will explain why AI bias will always exist—despite Trump’s executive order to ban “woke AI” and to sell AI embedding “American values” to U.S. allies. She will discuss how feminist theories like Hayles’s theory of cognitive assemblages may be correct but irrelevant in a world of digital oligarchs. And hopefully, we will end by discussing how, despite all this, we can continue to do useful research that contributes to a better world.

• – Date: Monday, 24 November 2025
  – Time: 12pm – 1:30pm (includes time for questions and discussion)
  – Location: Room 204.233 or online
  – Parking: Nearest parking is Pi1 or Pi2
  – Light refreshments provided; feel free to bring your lunch
  – RSVP: Please RSVP for in-person attendance to MCASIadmin@curtin.edu.au or join us online by scanning the QR Code on the 3rd page on the event day.

Sydney: Wednesday 26 Nov: AI-generated Australian stories: just like American stories but in the outback

Talk for the Literary Provoctions Hub at the University of New South Wales. Here is more info – it’s at 3:00pm-4:30pm at Robert Webster 327.

Abstract: What happens when you ask a large language model mostly trained on American training data to generate Australian stories? This talk uses recent developments in literary analysis of generative AI to analyse the homogeneity of AI-generated stories, with examples from a dataset of stories generated for each of 236 different nationalities. Almost all the generated stories share the same basic plot structure: a protagonist lives in or returns home to a small town and resolves a minor conflict by reconnecting with tradition and organising community events. A sprinkling of national flavour or stereotype is added – in the Australian stories the small town tends to be a mining town in the outback, the Norwegian towns are by the fjords – but the nostalgia, the lack of change and the absence of strong conflict or romance is constant across nationalities. We already know about AI bias: if you write “the doctor folded” the suggested next word is more likely to be “his”, and if you write “the nurse folded” it’ll probably be “her”. What if there is also a deeper bias, shaping narrative structure and not just the surface level of words? As generative AI is integrated into more and more of our writing tools, it is likely to nudge us into writing in particular ways. What happens if our computers nudge us to tell stories that are closer to Hollywood or Hallmark templates than to the stories of our own cultural contexts?

Canberra: Thursday 4 Dec: Archipelagos or Empires? Narrative Colonialism in Generative AI

Keynote at Digital Humanities Australasia 2025 (DHA2025) at Australian National University.

Abstract: What are the dominant narratives of generative AI, and what is at stake in their circulation? In this keynote, Jill Walker Rettberg discusses her ongoing research on AI-generated narratives in the AI STORIES project, which starts from the hypothesis that LLMs replicate and perhaps increase certain narrative patterns, which could mean that we lose diversity in storytelling. Research so far suggests this is true – the thousands of AI-generated stories we have analysed in the AI STORIES project emphasise stability and nostalgia, telling remarkably similar stories of threatened communities saved by reconnecting with heritage. The theme of DHA2025, Digital Archipelagos, reminds us both of our diversity and our interconnectedness – but can we retain these when using large language models (LLMs)? Is it possible to use large language models (LLMs) without succumbing to the digital colonialism of the large tech companies that sell them to us? How should we, as researchers and educators, respond to political and institutional pushes to use genAI? What does it mean for our digital archipelagos that Trump has issued an executive order banning “woke AI” and an AI Action Plan to ensure US allies use the “full AI technology stack” that aligns with American values? Generative AI is normalising, erasing the outliers and exceptions and replacing them with statistical probability. So would it help to use local models, or is the technology itself a problem? By understanding how LLMs really work we can gain the tools to decide when not to use it, and when it might add value. Rettberg will close by highlighting examples of how researchers might use LLMs with care, in ways that resist homogenisation and keep the archipelago alive.

AI STORIES team member Zahra Rizvi is also giving a talk at DHA2025:

Zahra Rizvi: Exploring Indian AI Stories (DHA2025 Canberra, 3 Dec, 1:30pm-3:00pm in the Grappling with AI session).

Abstract: Generative AI in India made its way into common people’s homes on Cadbury chocolates and Sunfeast biscuits, with Bollywood’s leading actor Shah Rukh Khan starring in AI generated ads for both the companies. The #NotJustACadburyAd, a collaboration between Cadbury, Shah Rukh Khan, and Mondelez India, aims to empower local store owners in India by utilizing AI and machine learning to create personalized advertisements featuring Shah Rukh Khan’s face and voice. The #MyFantasyAdWithSRK allowed fans to experience an AI generated experience of sharing Sunfeast biscuits with their favourite movie star. Both these ads were affected by the dominant narrative structures of marketing bylines and blurbs as well as the classist, almost Brahmanical, ideological systems that inform these narratives structures, from big-brand-ized ads for small businesses to the logic of exclusivity in the trope of the movie date.

The advent of large language models (LLMs) like GPT-4 has transformed text generation, offering innovative possibilities across various sectors, advertisement and marketing being just one area out of many. At the same time, as the example above illustrates, these models are subject to inherent biases, one of which is explored by the AI STORIES project (Rettberg 2024) as narrative bias. In my short paper, I will be describing the theoretical considerations of my ongoing postdoctoral work which posits that this bias emerges from the narrative archetypes embedded in the vast and culturally specific texts that LLMs are trained on, particularly those from English language sources, in my case, specifically Indian narratives. The implications of this bias can be profound, especially when one takes into account the context of cultural diversity and global storytelling practices from non-Western spaces like India.

This short paper will detail work in progress on how Indian narrative traditions, already rooted in a complex network of ancient epics, folktales, religious texts, and contemporary literature, are represented and potentially distorted by LLMs. In doing so, I intend to shed light on the transformation and mutation of culturally rich and diverse Indian narratives as they are processed by AI.