Julija Kalpokienė
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Julija Kalpokienė has a total of 9 co-authors. Here are the top co-authors by number of shared publications:
- Ignas Kalpokas
Vytautas Magnus University
15 shared publications - Donald Jay Bertulfo1 shared publication
- Enrico Panai
Università Cattolica del Sacro Cuore (UCSC), Milan, Italy
1 shared publication - Giada Pistilli
Hugging Face, France
1 shared publication - Leslye Denisse Dias Duran
Institute for Ethics in AI, University of Oxford, Oxford, United Kingdom
1 shared publication - Natalia Menéndez González
European University Institute
1 shared publication
Latest Publications
- 21 May 2026
- Published by Edward Elgar Publishing
- p. 350-363
Learning, agency, and Generative AI: a grassroots perspective
The use of Generative Artificial Intelligence (GenAI) in education is often framed as ‘cheating’ or ‘plagiarism’ while also raising fears of cognitive offloading, that is, the delegation of thinking and conceptualisation tasks to AI. Additional concerns pertain to errors in AI-generated content due to GenAI's limitations, particularly when used in smaller languages. To better understand the situation from a grassroots perspective, this chapter reports on a two-wave qualitative study that explores the experiences of MA students’ use of GenAI tools in one of their assignments. While uncovering a more interactive relationship between students and GenAI than much of the existing literature would suggest, implying that students retain significant agency in learning, this chapter also hints at deeper epistemic dependencies whereby not only the content but also the horizon of enquiry is tethered to the ‘black boxes’ of AI models, with a potential for tipping the balance of agency.
- 21 May 2026
- Published by Edward Elgar Publishing
- p. 350-363
Learning, agency, and Generative AI: a grassroots perspective
The use of Generative Artificial Intelligence (GenAI) in education is often framed as ‘cheating’ or ‘plagiarism’ while also raising fears of cognitive offloading, that is, the delegation of thinking and conceptualisation tasks to AI. Additional concerns pertain to errors in AI-generated content due to GenAI's limitations, particularly when used in smaller languages. To better understand the situation from a grassroots perspective, this chapter reports on a two-wave qualitative study that explores the experiences of MA students’ use of GenAI tools in one of their assignments. While uncovering a more interactive relationship between students and GenAI than much of the existing literature would suggest, implying that students retain significant agency in learning, this chapter also hints at deeper epistemic dependencies whereby not only the content but also the horizon of enquiry is tethered to the ‘black boxes’ of AI models, with a potential for tipping the balance of agency.
- Open Access
- 23 March 2026
- Published by Springer Nature in AI and Ethics
- Vol. 6 (2)
The ghost in the machine speaks with an American accent: cultural value drift in early GPT-3 and the case for pluralist evaluation of generative AI
Early large language models (LLMs) were released with minimal alignment, offering a rare view of how generative systems reframe the ethical values embedded in human texts. We examine outputs from a 2021 version of OpenAI’s base GPT-3, prompting it to summarise culturally diverse source materials (laws, political speeches, and philosophical works) and interpreting results through a descriptive, moral value pluralist lens. Where possible, we contextualise outputs with cross-national datasets such as the World Values Survey. We document recurring value drift: Australia’s firearm policy is recast as a threat to liberty; de Beauvoir’s feminist critique becomes gender-essentialist dating advice; and Merkel’s humanitarian appeal is recast as immigration control. In contrast, multilateral documents (UN/UNESCO) exhibit greater value stability, suggesting consensus-crafted language can buffer against cultural mutation. We argue that these early behaviours (observed before extensive fine-tuning and safety layers) provide a baseline for understanding how training distributions shape normative framing. Our contribution is twofold: (1) empirical evidence that value drift can invert or overwrite encoded values along predictable cultural axes, and (2) a pluralist, descriptive evaluation method that surfaces whose values dominate and when. We conclude with implications for culturally inclusive evaluation and alignment in contemporary LLMs.
- 23 March 2026
- Published by Springer Nature in AI and Ethics
- Vol. 6 (2)
Open AccessThe ghost in the machine speaks with an American accent: cultural value drift in early GPT-3 and the case for pluralist evaluation of generative AI
Early large language models (LLMs) were released with minimal alignment, offering a rare view of how generative systems reframe the ethical values embedded in human texts. We examine outputs from a 2021 version of OpenAI’s base GPT-3, prompting it to summarise culturally diverse source materials (laws, political speeches, and philosophical works) and interpreting results through a descriptive, moral value pluralist lens. Where possible, we contextualise outputs with cross-national datasets such as the World Values Survey. We document recurring value drift: Australia’s firearm policy is recast as a threat to liberty; de Beauvoir’s feminist critique becomes gender-essentialist dating advice; and Merkel’s humanitarian appeal is recast as immigration control. In contrast, multilateral documents (UN/UNESCO) exhibit greater value stability, suggesting consensus-crafted language can buffer against cultural mutation. We argue that these early behaviours (observed before extensive fine-tuning and safety layers) provide a baseline for understanding how training distributions shape normative framing. Our contribution is twofold: (1) empirical evidence that value drift can invert or overwrite encoded values along predictable cultural axes, and (2) a pluralist, descriptive evaluation method that surfaces whose values dominate and when. We conclude with implications for culturally inclusive evaluation and alignment in contemporary LLMs.
- 3 March 2026
- Published by Taylor & Francis
- p. 148-163
I VR therefore I am: toxic binary thinking in visions of the metaverse
While the metaverse is typically billed as representing a radical break with the past and an invention of a new reality in a virtual form it merely reproduces historical patterns of anthropocentric dominance, exploitation, and violence. The Western intellectual tradition has long billed the human (or a privileged part of humanity) as the bearer of special rights, moral status, and epistemic capacities, which were seen as rendering the rest of the world not only subservient, but even existing solely for human benefit. Moreover, particularly since the Enlightenment, this dualism was also extended within the human by drawing a dividing line between the body and the mind. To trace the morphing of such ideas into the digital domain, this article provides a critical reading of the mainstream accounts of the metaverse. Crucially, the metaverse functions as a perfect manifestation of the vision for a domain of the mind, radically separated from all embodiment and always adjusting to human needs through data-based rendering of virtual worlds. However, the metaverse not only represents the conquest of the embodied physical world but also reproduces expansionist and colonialist tropes of (virtual) land grab, subjection of the world to human will, and exploitation of (digital) resources. In this way, visions for the metaverse enable seemingly guilt-free reenactment of the behaviours and attitudes that have led to great injustices in the past and in the present.
- 3 March 2026
- Published by Taylor & Francis
- p. 148-163
I VR therefore I am: toxic binary thinking in visions of the metaverse
While the metaverse is typically billed as representing a radical break with the past and an invention of a new reality in a virtual form it merely reproduces historical patterns of anthropocentric dominance, exploitation, and violence. The Western intellectual tradition has long billed the human (or a privileged part of humanity) as the bearer of special rights, moral status, and epistemic capacities, which were seen as rendering the rest of the world not only subservient, but even existing solely for human benefit. Moreover, particularly since the Enlightenment, this dualism was also extended within the human by drawing a dividing line between the body and the mind. To trace the morphing of such ideas into the digital domain, this article provides a critical reading of the mainstream accounts of the metaverse. Crucially, the metaverse functions as a perfect manifestation of the vision for a domain of the mind, radically separated from all embodiment and always adjusting to human needs through data-based rendering of virtual worlds. However, the metaverse not only represents the conquest of the embodied physical world but also reproduces expansionist and colonialist tropes of (virtual) land grab, subjection of the world to human will, and exploitation of (digital) resources. In this way, visions for the metaverse enable seemingly guilt-free reenactment of the behaviours and attitudes that have led to great injustices in the past and in the present.
- 30 October 2025
- Published by Taylor & Francis
- p. 93-109
From Robot Wars to Space Colonialism
This chapter examines the perceived imperative to use technological capacities for maximising ethically questionable outcomes, focusing on lethal autonomous weapons systems (LAWS) and space exploitation. In warfare, LAWS promise speed and efficiency but risk deepening inequalities, allowing technologically advanced forces to avoid harm while opponents bear human costs. By removing humans from critical decisions, these systems can dehumanise conflict and detach killing from ethical responsibility. In space, ambitions for colonisation mirror historical colonial ideologies, treating extraterrestrial environments as resources for extraction while neglecting ethical considerations, and are demonstrated to be riddled with ethical issues. Moreover, such ventures risk perpetuating social inequities and only further exacerbate ongoing problems that space colonialism is said to help avoid (e.g., climate change). Both domains reflect a mindset where technological capability becomes a mandate for action, turning ‘can’ into ‘must’ and sidelining moral deliberation. The chapter challenges this deterministic narrative, emphasising the need to recognise and address the ethical crossroads before pursuing technological paths that could further contribute to inequality and injustice.
- 30 October 2025
- Published by Taylor & Francis
- p. 93-109
From Robot Wars to Space Colonialism
This chapter examines the perceived imperative to use technological capacities for maximising ethically questionable outcomes, focusing on lethal autonomous weapons systems (LAWS) and space exploitation. In warfare, LAWS promise speed and efficiency but risk deepening inequalities, allowing technologically advanced forces to avoid harm while opponents bear human costs. By removing humans from critical decisions, these systems can dehumanise conflict and detach killing from ethical responsibility. In space, ambitions for colonisation mirror historical colonial ideologies, treating extraterrestrial environments as resources for extraction while neglecting ethical considerations, and are demonstrated to be riddled with ethical issues. Moreover, such ventures risk perpetuating social inequities and only further exacerbate ongoing problems that space colonialism is said to help avoid (e.g., climate change). Both domains reflect a mindset where technological capability becomes a mandate for action, turning ‘can’ into ‘must’ and sidelining moral deliberation. The chapter challenges this deterministic narrative, emphasising the need to recognise and address the ethical crossroads before pursuing technological paths that could further contribute to inequality and injustice.
- 30 October 2025
- Published by Taylor & Francis
- p. 15-29
Solutionist Imaginaries
- 30 October 2025
- Published by Taylor & Francis
- p. 110-114
Conclusions
Technological progress is increasingly framed as inevitable and universally beneficial, sidelining moral deliberation. The failure to recognise crossroads in decision-making – points where alternative paths and values could be considered – creates a moral vacuum, or blindness, in which efficiency and optimisation override ethical concerns. Across domains from education and creativity to warfare, space colonisation, and transhumanism, technological possibility becomes an imperative. This frictionless trajectory privileges dominant interests, marginalises alternative visions, and normalises exploitation, making resistance or even noticing alternatives ever more difficult.
- 30 October 2025
- Published by Taylor & Francis
- p. 110-114
Conclusions
Technological progress is increasingly framed as inevitable and universally beneficial, sidelining moral deliberation. The failure to recognise crossroads in decision-making – points where alternative paths and values could be considered – creates a moral vacuum, or blindness, in which efficiency and optimisation override ethical concerns. Across domains from education and creativity to warfare, space colonisation, and transhumanism, technological possibility becomes an imperative. This frictionless trajectory privileges dominant interests, marginalises alternative visions, and normalises exploitation, making resistance or even noticing alternatives ever more difficult.
- 30 October 2025
- Published by Taylor & Francis
- p. 58-74
Towards the (Last) Frontier
This chapter examines the growing influence of virtual persons and their impact on human interactions and consumer behaviour. The first part is dedicated to virtual influencers (VIs), created and controlled by brands or third parties, that increasingly carve their own niche alongside more ‘traditional’ human counterparts. Further on, additional types of virtual persons are introduced, such as avatars or even more or less independent incarnations of real people, as well as virtual companions. All of the preceding raise ethical concerns, including potential for audience manipulation, the creation of unrealistic beauty standards, the likelihood of fostering toxic relationship dynamics, the possibility for creation of highly normative designer personalities, and the devaluation of human uniqueness in a future increasingly dominated by digital representations. Ultimately, virtual personalities are seen as a significant cultural force, raising important questions about their influence on identity, authenticity, and societal norms.
- 30 October 2025
- Published by Taylor & Francis
- p. 58-74
Towards the (Last) Frontier
This chapter examines the growing influence of virtual persons and their impact on human interactions and consumer behaviour. The first part is dedicated to virtual influencers (VIs), created and controlled by brands or third parties, that increasingly carve their own niche alongside more ‘traditional’ human counterparts. Further on, additional types of virtual persons are introduced, such as avatars or even more or less independent incarnations of real people, as well as virtual companions. All of the preceding raise ethical concerns, including potential for audience manipulation, the creation of unrealistic beauty standards, the likelihood of fostering toxic relationship dynamics, the possibility for creation of highly normative designer personalities, and the devaluation of human uniqueness in a future increasingly dominated by digital representations. Ultimately, virtual personalities are seen as a significant cultural force, raising important questions about their influence on identity, authenticity, and societal norms.
- 30 October 2025
- Published by Taylor & Francis
- p. 1-14
A One-Way Street or Crossroads
It is a truism that technology, such as AI, increasingly pervades every aspect of contemporary societies. This techno-centricity, however, is neither neutral nor inevitable but, instead, shaped by human choices and values, interwoven with human intentions, cultural norms, and societal structures. It reflects and amplifies existing inequalities, biases, and ideologies while introducing new dynamics of power and control. And yet, the way in which technology, and the broader notion of progress, is framed not only obfuscates such contingencies and value-based choices – it presents technological development as both unavoidable and a good in itself. Potential crossroads in this path of progress not only fail to be seen – their very possibility is denied. Hence, this chapter explores the interplay between technological progress and societal values, emphasising the need for critical reflection on the trajectories of digital advancements. It challenges the deterministic narrative that equates technological development with linear progress, highlighting the moral blindness and ethical oversights that arise from this perspective. Drawing on the concepts of adiaphorisation and liquid evil, it challenges the supposed consensus around technological progress as unequivocal good, emphasising the need to notice and seriously consider the value crossroads of technological advancements.
- 30 October 2025
- Published by Taylor & Francis
- p. 1-14
A One-Way Street or Crossroads
It is a truism that technology, such as AI, increasingly pervades every aspect of contemporary societies. This techno-centricity, however, is neither neutral nor inevitable but, instead, shaped by human choices and values, interwoven with human intentions, cultural norms, and societal structures. It reflects and amplifies existing inequalities, biases, and ideologies while introducing new dynamics of power and control. And yet, the way in which technology, and the broader notion of progress, is framed not only obfuscates such contingencies and value-based choices – it presents technological development as both unavoidable and a good in itself. Potential crossroads in this path of progress not only fail to be seen – their very possibility is denied. Hence, this chapter explores the interplay between technological progress and societal values, emphasising the need for critical reflection on the trajectories of digital advancements. It challenges the deterministic narrative that equates technological development with linear progress, highlighting the moral blindness and ethical oversights that arise from this perspective. Drawing on the concepts of adiaphorisation and liquid evil, it challenges the supposed consensus around technological progress as unequivocal good, emphasising the need to notice and seriously consider the value crossroads of technological advancements.
- 30 October 2025
- Published by Taylor & Francis
- p. 30-44
(Re)Locating the Author
While AI and associated technologies have long been seen as threatening manual, repetitive, and easily predictable jobs, creative occupations were seen as relatively immune to automation. However, the emergence of generative AI has thrust the previously niche subject of AI creativity into the public debate. At the very least, generative AI can be seen as optimising artistic endeavour and maximising the efficiency of output production – in essence, putting creativity on the same page as any other domains about to be impacted by AI, that is, subjecting creativity to optimising logic. However, AI’s capacity for shrinking the time for learning and creation, as well as to learn from and emulate existing examples of human creativity, means that, at least in theory, replacement of humans in creative endeavours is possible. Perhaps even more fundamentally, though, AI has the capacity to separate creativity and human agency, thereby undermining historically taken-for-granted assumptions. In this way, crossroads are opened pertaining to the role of humans and potential lines of demarcation in artistic and creative contexts.
- 30 October 2025
- Published by Taylor & Francis
- p. 30-44
(Re)Locating the Author
While AI and associated technologies have long been seen as threatening manual, repetitive, and easily predictable jobs, creative occupations were seen as relatively immune to automation. However, the emergence of generative AI has thrust the previously niche subject of AI creativity into the public debate. At the very least, generative AI can be seen as optimising artistic endeavour and maximising the efficiency of output production – in essence, putting creativity on the same page as any other domains about to be impacted by AI, that is, subjecting creativity to optimising logic. However, AI’s capacity for shrinking the time for learning and creation, as well as to learn from and emulate existing examples of human creativity, means that, at least in theory, replacement of humans in creative endeavours is possible. Perhaps even more fundamentally, though, AI has the capacity to separate creativity and human agency, thereby undermining historically taken-for-granted assumptions. In this way, crossroads are opened pertaining to the role of humans and potential lines of demarcation in artistic and creative contexts.
- 30 October 2025
- Published by Taylor & Francis
- p. 45-57
Many Problems, Few Solutions
This chapter explores the questions as to what counts as digital heritage, what should be protected and by whom, and how the decisions on selecting heritage-to-be ought to be made. Arguably, such questions have become more acute and complex than ever before, not only due to the unprecedented quantity and variety of content but also due to the rise of new actors and intermediaries. Simultaneously, the centrality of digital content in the everyday lives of individuals means that protection of digital content – or at least of a representative sample thereof – is a must if an adequate record of contemporary life is to be preserved for posterity. The focus of this chapter is on born-digital content that originates from online interactions or is used to host and structure such interactions (e.g., virtual environments), that is, mundane heritage as everyday practices of today’s digital-by-default life. Here, contemporary societies are again at a crossroads, facing crucial choices regarding the role and purpose of AI in shaping not just the present but also the future.
- 30 October 2025
- Published by Taylor & Francis
- p. 45-57
Many Problems, Few Solutions
This chapter explores the questions as to what counts as digital heritage, what should be protected and by whom, and how the decisions on selecting heritage-to-be ought to be made. Arguably, such questions have become more acute and complex than ever before, not only due to the unprecedented quantity and variety of content but also due to the rise of new actors and intermediaries. Simultaneously, the centrality of digital content in the everyday lives of individuals means that protection of digital content – or at least of a representative sample thereof – is a must if an adequate record of contemporary life is to be preserved for posterity. The focus of this chapter is on born-digital content that originates from online interactions or is used to host and structure such interactions (e.g., virtual environments), that is, mundane heritage as everyday practices of today’s digital-by-default life. Here, contemporary societies are again at a crossroads, facing crucial choices regarding the role and purpose of AI in shaping not just the present but also the future.