Forskning og undersøgelser

Elevers brug af AI i gymnasiet

I denne undersøgelse kan du blive klogere på, hvordan gymnasieelever benytter sig af AI i forbindelse med skolearbejdet. Hvordan oplever de, at AI påvirker deres læring? Hvad ansporer dem til at anvende AI? Og hvordan vurderer de den undervisning i AI, de får på gymnasiet i 2024/25?

Elevers brug af AI i gymnasiet –  Tilgange og holdninger til brug af AI i undervisning og skolearbejde


Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task

Kosmyna, Nataliya et al ;10.48550/arXiv.2506.08872  – arXiv:arXiv:2506.08872

Publication:  eprint arXiv:2506.08872; Pub Date:June 2025. DOI:

E-Print Comments: 216 pages, 102 figures, 4 tables and appendix

This study explores the neural and behavioral consequences of LLM-assisted essay writing.

As the educational impact of LLM use only begins to settle with the general population, in this study we demonstrate the pressing matter of a likely decrease in learning skills based on the results of our study. The use of LLM had a measurable impact on participants, and while the benefits were initially apparent, as we demonstrated over the course of 4 sessions, which took place over 4 months, the LLM group’s participants performed worse than their counterparts in the Brain-only group at all levels: neural, linguistic, scoring

We used electroencephalography (EEG) to assess cognitive load during essay writing, and analyzed essays using NLP, as well as scoring essays with the help from human teachers and an AI judge. EEG revealed significant differences in brain connectivity: Brain-only participants exhibited the strongest, most distributed networks; Search Engine users showed moderate engagement; and LLM users displayed the weakest connectivity. Self-reported ownership of essays was the lowest in the LLM group and the highest in the Brain-only group. LLM users also struggled to accurately quote their own work.

While LLMs offer immediate convenience, our findings highlight potential cognitive costs. Over four months, LLM users consistently underperformed at neural, linguistic, and behavioral levels.

EEG analysis presented robust evidence that LLM, Search Engine and Brain-only groups had significantly different neural connectivity patterns, reflecting divergent cognitive strategies. Brain connectivity systematically scaled down with the amount of external support: the Brain‑only group exhibited the strongest, widest‑ranging networks, Search Engine group showed intermediate engagement, and LLM assistance elicited the weakest overall coupling. In session 4, LLM-to-Brain participants showed weaker neural connectivity and under-engagement of alpha and beta networks; and the Brain-to-LLM participants demonstrated higher memory recall, and re‑engagement of widespread occipito-parietal and prefrontal nodes, likely supporting the visual processing, similar to the one frequently perceived in the Search Engine group. The reported ownership of LLM group’s essays in the interviews was low. The Search Engine group had strong ownership, but lesser than the Brain-only group. The LLM group also fell behind in their ability to quote from the essays they wrote just minutes prior.


Is It Harmful or Helpful? Examining the Causes and Consequences of Generative AI Usage among University Students.

Abbas, M., F. A. Jam, and T. I. Khan. 2024. International Journal of Educational Technology. https://doi.org/10.1186/s41239-024-00444-7

The use of ChatGPT among students and the potential harmful or beneficial consequences associated with its usage.

Keywords: Workload, Time pressure, Sensitivity to quality, Sensitivity to rewards, ChatGPT usage, Procrastination, Memory loss, Academic performance

.. Using samples from two studies, the current research examined the causes and consequences of ChatGPT usage among university students.

Study 1 developed and validated an eight-item scale to measure ChatGPT usage by conducting a survey among university students. Study 2 used a three-wave time-lagged design to collect data from university students to further validate the scale and test the study’s hypotheses. Study 2 also examined the effects of academic workload, academic time pressure, sensitivity to rewards, and sensitivity to quality on ChatGPT usage. Study 2 further examined the effects of ChatGPT usage on students’ levels of procrastination, memory loss, and academic performance.

Study 1 provided evidence for the validity and reliability of the ChatGPT usage scale. Furthermore, study 2 revealed that when students faced higher academic workload and time pressure, they were more likely to use ChatGPT. In contrast, students who were sensitive to rewards were less likely to use ChatGPT.

Not surprisingly, use of ChatGPT was likely to develop tendencies for procrastination and memory loss and dampen the students’ academic performance. Finally, academic workload, time pressure, and sensitivity to rewards had indirect effects on students’ outcomes through ChatGPT usage.


The Usage of New AI Technologies by College Students and Its Influence on Learning and Dependence.

Ayele, Nathan (2024). Honors Projects: Open Access (2019–present). Caldwell University Archives. https://www.jstor.org/stable/community.37312334

Hvad der sker på den korte bane ved brug af disse værktøjer, særligt med kritisk tænkning og evnen til at reflektere?

Survey of 61 students found a double-edged sword: while students gained efficiency, many reported a concerning decrease in confidence in their academic skills and more dependence on these tools. Utility and Efficiency of AI vs. Dependence on AI: A common concern among students is the over-reliance on AI technologies. They worry that this dependence might make them lazy and less confident in their academic skills.

A few students expressed,

“It was made to help us and now it’s just taking over us…”

”I have noticed that although it makes things faster for me by formulating my thoughts and information I give it in a proper way, it is taking away from my own ability to practice my writing on my own.”


AI Tools in Society: Impacts on Cognitive Offloading and the Future of Critical Thinking

Gerlich, Michael. 2025. Societies 15, no. 1: 6. https://doi.org/10.3390/soc15010006


Impact of technostress on work-life balance.

Bencsik, A., & Juhasz, T. (2023). Human Technology, 19(1), 41–61. https://doi.org/10.14254/1795-6889.2023.19-1.4

The aim of this research is to identify the most  significant  risks  factors  of  technostress  that  threaten  the  balanced performance  of employees at work and the possibility of work-life balance. 

The most noticeable effects of  technostress that can be observed in the functioning of organisations are: Increased role overload; Role conflict; Reduced job satisfaction; Fatigue and burnout. 

A proprietary model was built to test the impact of the most serious risks on personal and work  life. The results show that three factors have the  greatest impact on work-life balance, which also affect organisational performance. Loss  of  leisure  time  due  to techno-overload and techno-invasion, and a feeling of techno-uncertainty due to lack of ICT competence cause a sense of threat.

These compromise work-life balance and, at the same time, work performance. The mutually reinforcing negative effects influence the sense of well-being (happiness) at work, the feeling of job security and force employees to learn continuously. In fact, learning new  skills  requires  more  effort, may initially slow down work processes, and the fact that additive effort is often only  possible beyond working hours and not within them is an additional stress factor.  This  increases the problem of reconciling work and private life and poses a risk to performance at work.

 


Artificial Intelligence and the Real Existential Risks: An Analysis of the Human Limitations of Control.

Birolo Candiotto, Kleber Bez, and Murilo Karasinski. 2022.  Filosofia Unisinos 23, (3). https://doi.org/10.4013/fsu.2022.233.07

Keywords: Philosophy, human supremacy, biological brain, uncritical thinking.    

Based on the hypothesis that artificial intelligence would not represent the end of human supremacy, since, in essence, AI would only simulate and increase aspects of human intelligence in non-biological artifacts, this paper questions the real risk to be faced. Beyond the clash between technophobes and technophiles, what is argued, then, is that the possible malfunctions of an artificial intelligence – resulting from information overload, from a wrong programming or from a randomness of the system – could signal the real existential risks, especially when we consider that the biological brain, in the wake of the automation bias, tends to assume uncritically what is set by systems anchored in artificial intelligence.

Moreover, the argument defended here is that failures undetectable by the probable limitation of human control regarding the increased complexity of the functioning of AI systems represent the main real existential risk.


Humans versus Machines. In The AI Book

Blomstrom, D. (2020). (eds S. Chishti, I. Bartoletti, A. Leslie and S.M. Millie). https://doi-org.ep.fjernadgang.kb.dk/10.1002/9781119551966.ch72


Implications of AI for Work, Employment and Social Dialogue: Literature Review.

Bugajska, Anna. 2024. Forum philosophicum (Kraków, Poland), 2024-12, Vol.29 (2), p.351-369


Hope Beyond Human: A Philosophy of Hope in the Digital Age

 


The Emotional Risk Posed by AI (Artificial Intelligence) in the Workplace.

Danielsen, M. (2023). Norsk filosofisk tidsskrift58(2–3), 106–117. https://doi.org/10.18261/nft.58.2-3.4

Keywords: Emotional risk from Ai. Loss (of care and meaning). Ai at work.

The existential risk posed by ubiquitous artificial intelligence (AI) is a subject of frequent discussion with descriptions of the prospect of misuse, the fear of mass destruction, and the singularity.

In this paper I address an underexplored category of existential risk posed by AI, namely emotional risk. Values are a main source of emotions. By challenging some of our most essential values, AI systems are therefore likely to expose us to emotional risks such as loss of care and loss of meaning.

Part one presents a study of a leading bank in Germany where an AI system was implemented to replace humans in decision-making processes. Part two explains why humans actively make use of values to make decisions. Part three shows the connection between values and emotions. Part four relates parts two and three to the bank study by giving concrete examples of how the employees saw their roles as workers, and how the relationships to their customers changed emotionally after the AI system was implemented to make decisions.


 The Cybernetic Teammate: A Field Experiment on Generative AI Reshaping Teamwork and Expertise. 

Dell’Acqua, Fabrizio, Charles Ayoubi, Hila Lifshitz, Raffaella Sadun, Ethan Mollick, Lilach Mollick, Yi Han, Jeff Goldman, Hari Nair, Stew Taub, and Karim R. Lakhani. Harvard Business School Working Paper, No. 25-043, March 2025. https://www.hbs.edu/faculty/Pages/item.aspx?num=67197

Generative AI Reshaping Teamwork and Expertise

 


AI & Human Values.

Devillers, L., Fogelman-Soulié, F., Baeza-Yates, R. (2021). In: Braunschweig, B., Ghallab, M. (eds) Reflections on Artificial Intelligence for Humanity. Lecture Notes in Computer Science(), vol 12600. Springer, Cham. https://doi-org.kb-ku.idm.oclc.org/10.1007/978-3-030-69128-8_6

This chapter summarizes contributions made by Ricardo Baeza-Yates, Francesco Bonchi, Kate Crawford, Laurence Devillers and Eric Salobir in the session chaired by Françoise Fogelman-Soulié on AI & Human values at the Global Forum on AI for Humanity.

It provides an overview of key concepts and definitions relevant for the study of inequalities and Artificial Intelligence. It then presents and discusses concrete examples of inequalities produced by AI systems, highlighting their variety and potential harmfulness. Finally, we conclude by discussing how putting human values at the core of AI requires answering many questions, still open for further research.

 


Generative AI Enhances Individual Creativity but Reduces the Collective Diversity of Novel Content.

Doshi, Anil K., and Oliver Hauser. 2024. Science Advances 10 (28): DOI:10.1126/sciadv.adn5290

Creativity is core to being human. Generative artificial intelligence (AI)—including powerful large language models (LLMs)—holds promise for humans to be more creative by offering new ideas, or less creative by anchoring on generative AI ideas.

We study the causal impact of generative AI ideas on the production of short stories in an online experiment where some writers obtained story ideas from an LLM. We find that access to generative AI ideas causes stories to be evaluated as more creative, better written, and more enjoyable, especially among less creative writers.

However, generative AI–enabled stories are more similar to each other than stories by humans alone. These results point to an increase in individual creativity at the risk of losing collective novelty. This dynamic resembles a social dilemma: With generative AI, writers are individually better off, but collectively a narrower scope of novel content is produced.

Our results have implications for researchers, policy-makers, and practitioners interested in bolstering creativity.

 


The Empowerment of Artificial Intelligence in Post‑Digital Organisations: Exploring Human Interactions with Supervisory AI.

Gladden, Matthew, Paweł Fortuna, and Artur Modliński. 2022. Human Technology 18 (2): 98–121. https://ht.csr-pub.eu/index.php/ht

Organisationers brug af teknologi og hvordan den bruges i dag og et kig ind i fremtiden.       

Technology evolves together with humans. Across industrial revolutions, its role has evolved from that of a simple tool used by humans to that of intelligent decision-maker and teammate. In the post-digital era where ongoing advances in artificial intelligence are widely visible, the  question  arises  regarding  the  extent  to  which  technology  will  be “upgraded” into roles previously filled by human supervisors, thereby replacing persons in  managerial  positions. 

This text  aims  to  delineate  how  the  organizational  role  of technology  has  been  transformed  across  decades  and  the  forms  that it  currently  takes within companies, with an eye to the future.

We draw on posthuman managerial literature and known cases of organizations where some forms of supervisory artificial intelligence are  already  used.  The  text  is  conceptual-reflective  by  nature;  it  seeks  to  initiate  a discussion on the many  challenges  that  humanity  will  face  in  connection  with  the deployment of empowered posthuman agents in companies.