Algorithmic Assistance and the Creativity Paradox: Large Language Models, Student Writing, and Cognitive Homogenisation

The adoption of large language models in higher education aids brainstorming and productivity, but raises concerns about student dependence, potentially diminishing confidence in generating original ideas independently. The paper tackles these issues by carrying out a quantitative survey on a sample size of 120 undergraduate and postgraduate students of four universities in Dublin, that is, Dublin City University, University College Dublin, Trinity College Dublin, and Technological University Dublin. The respondents were varied in their academic backgrounds; some were undergraduate or post-graduate level and this was evenly split between 46 per cent female and 54 per cent male. The survey was taking the trends of using ChatGPT, the sense of creative self-efficacy, and how dissimilar students perceived academic ideas and written work. The results suggest that there is a definite conflict in the application of AI tools. A significant number of respondents referred to ChatGPT in the context of the idea development, increased productive, and assistance at the first drafting phase. Nevertheless, a large percentage of them experienced more trouble in idea generation, and thought they were beginning to resemble the work of other students using the same system. On the whole, 67.5 per cent thought that their ideas were growing closer to those who also use ChatGPT. These results imply that the positive effects of writing with AI might be accompanied by the feeling of reduced originality. The paper highlights AI’s impact on creativity, emphasizing pedagogical strategies for fostering independent thinking and intellectual diversity in education.