From Print to Pixels to Prompts: Stage-Switching as a Critical Competence for AI-Mediated E-Learning

Abstract

This perspective article argues that digital reading within e-learning environments should not be conceptualized as a static shift from print to screen, but as an evolving continuum spanning three overlapping modalities: print reading, e-reading (pixels), and prompt-mediated reading (prompts). In this third stage, generative AI systems dynamically synthesize, curate, or interpret text on behalf of the learner. Drawing on digital literacy frameworks, cognitive load theory, and recent scholarship on generative AI in education, this paper introduces a conceptual model of this reading continuum. It systematically examines the pedagogical affordances and cognitive trade-offs of each stage regarding comprehension, sustained attention, and learner autonomy. Rather than framing AI-mediated text engagement as an academic threat to be restricted or an uncritical tool to be adopted passively, this article calls for an explicit curricular reframing around stage-switching the deliberate pedagogical capability to navigate across these modalities. The paper concludes by outlining concrete implications for e-learning instructional designers, curriculum developers, and future empirical research into digital reading behaviors within AI-saturated educational ecosystems.

Introduction

The text argues that the concept of “doing the reading” in e-learning has changed because students now interact with texts not only through print and digital screens but also through generative AI. Traditional research often treats reading as a simple print-versus-digital distinction, but the emergence of AI-assisted reading creates a third, fundamentally different practice.

The paper proposes the “Print–Pixels–Prompts Continuum”:

• Print: Linear, physically anchored reading that supports sustained attention, deep comprehension, and spatial memory. Its limitations include slower information retrieval and limited connectivity.

• Pixels: Screen-based reading involving hyperlinks, searching, skimming, and navigating multiple sources. It increases information access and supports hyper-reading, but can also produce distraction and fragmented attention.

• Prompts: AI-mediated reading in which learners ask an LLM to summarize, compare, clarify, or analyze texts. This provides rapid synthesis and information triage but introduces risks such as over-reliance on summaries, hallucinations, algorithmic bias, and reduced engagement with the original text.

A central argument is that prompt-mediated reading is not simply a faster form of digital reading. It changes the learner’s role from directly interpreting a source to directing and evaluating an algorithmic interpretation of that source. Therefore, students need new forms of critical literacy, including the ability to verify AI outputs, detect hallucinations and omissions, recognize bias, and compare AI-generated interpretations with primary sources.

The paper emphasizes that the three modalities do not replace one another. Students may move recursively between them—for example, using an AI tool for an initial overview, examining a PDF for detailed evidence, and then returning to print for close annotation. The authors call the ability to move deliberately between these modes “stage-switching,” and propose that it should become an important metacognitive skill in contemporary education.

Conclusion

The contemporary debate surrounding generative AI in education has suffered from a profound conceptual bottleneck. By attempting to force prompt-mediated textual engagement into the legacy paradigms of the print-versus-digital divide, educational technology research has overlooked a fundamental shift in learner cognition. This perspective article has introduced the Print–Pixels–Prompts continuum to provide a precise, operationally viable lexicon for this new landscape. Prompt-mediated reading represents a clean break from traditional digital text interaction; it shifts the educational task from human information processing to human-algorithmic orchestration. As e-learning environments become increasingly saturated with agentic AI systems, the preservation of autonomous learning requires an immediate shift in empirical research priorities. Crucial questions must now be systematically addressed by the educational technology community: 1) What are the measurable developmental impacts of protracted prompt-mediated reading habits on the baseline long-form reading comprehension capacities of adolescent learners over extended academic cycles? 2) Does explicit instructional training in metacognitive stage-switching yield quantifiable improvements in critical thinking and source verification metrics compared to unguided technology adoption? 3) What do high-competence, rigorous forms of prompt-mediated reading look like in practice, and how can these expert strategies be operationalized into scalable instructional scaffolds within digital course designs? The answers to these inquiries will determine whether generative AI serves as a mechanism for widespread cognitive offloading, or as a catalyst for a sophisticated new tier of human-machine intellectual partnership. What is certain is that the baseline definition of what it means to \"do the reading\" has shifted permanently. Educational systems must either intentionally adapt their pedagogical architectures to manage this continuum or leave learners to navigate its structural risks by chance.

References

[1] Hakemulder, F., & Mangen, A. (2024). Literary Reading on Paper and Screens: Associations Between Reading Habits and Preferences and Experiencing Meaningfulness. Reading Research Quarterly, 59(1). https://doi.org/10.1002/rrq.527 [2] Hayles, N. K. (2025). Modes of Cognition: Implications for Large Language Models. Antikythera Digital Journal. https://doi.org/10.1162/anti.5czf [3] Lucciarini, E., Carr, A., Bressoud, N., Boniwell, I., Audrin, C., Gay, P., & Burel, N. (2025). Correction to: Evermore Wellbeing for Teens: A Scoping Review of Stand-alone and Brief Positive Psychology Interventions for Potential Use in Educational Settings (Educational Psychology Review, (2025), 37, 3, (63), 10.1007/s10648-025-10033-3). In Educational Psychology Review (Vol. 37, Issue 3). https://doi.org/10.1007/s10648-025-10064-w [4] Mangen, A., Olivier, G., & Velay, J. L. (2019). Comparing comprehension of a long text read in print book and on kindle: Where in the text and when in the story? Frontiers in Psychology, 10(FEB). https://doi.org/10.3389/fpsyg.2019.00038 [5] Ruffini, C., Delgado, P., Saldaña, D., & Pecini, C. (2025). Third graders’ digital and paper text comprehension. Learning and Instruction, 100. https://doi.org/10.1016/j.learninstruc.2025.102186 [6] Wang, S., Guo, C., Yu, T., Wang, K., Wei, Q., & He, Y. (2026). From attentional bias to neural dysfunction: Multimodal evidence for emotion processing deficits in individuals with problematic social media use. Journal of Behavioral Addictions. https://doi.org/10.1556/2006.2025.00306

Copyright

Copyright © 2026 Sahaya James, B R Veeramani James. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.