Tom Colloton

The rapid integration of large language models (LLMs) into education has shifted research focus from questions of capability, such as what LLMs can do and how accurately—to questions of deployability, including how they can be operated effectively for many learners at once. In school environments, system reliability, scalability, and real-time responsiveness are critical, as delays or interruptions can directly reduce learner engagement, particularly during synchronous activities. This study evaluates the performance of open-source LLaMA models ranging from 1 billion to 70 billion parameters across one-, dual-, triple-, and quad-GPU configurations suitable for educational settings. Performance is assessed using four key indicators: success rate (percentage of completed requests), generation speed (tokens per second), throughput (completed responses per second), and latency (time until full response generation). These metrics were measured under progressively increasing numbers of simultaneous users to identify system capacity limits and trade-offs between model size, responsiveness, and scalability. The results indicate that smaller models (1B–3B) deliver faster, more stable performance under concurrent use, while larger models (8B–70B) experience substantial slowdowns and reduced reliability, even on high-end GPU systems. These findings suggest that effective educational deployment should prioritize empirical performance and infrastructure compatibility over model size alone. The paper concludes by proposing a practical framework to guide educators, administrators, and developers in selecting and configuring locally hosted GPU systems that balance model capability, response speed, and resource efficiency for real-time applications such as AI tutors, classroom chatbots, and automated feedback tools.

Opening with an introductory backdrop on Artificial Intelligence (AI), Big Data, the Internet of Things (IoT) and Generative AI (GenAI), this chapter offers readers foundational knowledge of these groundbreaking technologies and their specific applications within educational settings. The spotlight is cast on ChatGPT by OpenAI, highlighting its potential as a revolutionary tool in higher education. A detailed breakdown of AI, from its inception in the 1950s to its current categorisations – Artificial Narrow Intelligence (ANI), Artificial General Intelligence (AGI), and Artificial Super Intelligence (ASI) – is provided. The chapter then delves into the significance of Big Data, IoT and GenAI and their intrinsic characteristics, with a focused example on their implications for higher education. The chapter connects technological advancements with their practical implications in education, urging readers to consider the potential and pitfalls of integrating AI in the classroom.

This chapter provides an in-depth exploration of ChatGPT and GenAI’s implications in higher education curriculum design. The chapter underscores the strengths, potential opportunities, weaknesses, and threats posed by GenAI in higher education settings. Drawing from a wealth of resources including literature reviews, online discussions, and firsthand testimonials from students and educators, the chapter offers a list of practical examples of how ChatGPT can be woven into pedagogical design. Serving as a roadmap, the list encourages stakeholders to merge AI capabilities with human ingenuity to foster innovative teaching methods. In addition, the chapter demonstrates a research study examining the synergy between ChatGPT and Bloom’s Taxonomy to present a holistic view of AI-partnered pedagogy. The chapter concludes with insights into prompt engineering and its components, offering guidance on optimising interactions with GenAI.

Chapter 2 of the book emphasises the crucial nature of AI literacy in today’s technologically-driven society and provides a perceptive definition based on literature and research findings. It underscores the necessity of possessing foundational AI literacy, which grants individuals the essential understanding to navigate the modern digital landscape, regardless of their profession. However, the chapter posits that AI literacy cannot remain broad-brushed; it needs further granularity based on professional contexts. For instance, while healthcare professionals may grapple with life-threatening AI mistakes, business professionals might harness AI to shape brand perceptions. Consequently, the Dynamic AI Literacy Model (DAILM) for Specific Roles is introduced, proposing a layered approach to AI literacy tailored to specific professions. Additionally, the chapter sheds light on AI literacy’s importance in the broader digital literacy spectrum, with a special focus on AI literacy for teachers, and also evaluates various AI literacy frameworks from previous research.

Chapter 5 delves into formulating policies for Artificial Intelligence (AI) in education, emphasising the importance of understanding regulatory principles like transparency, fairness, accountability, safety, and privacy across platforms such as ChatGPT and GenAI. The chapter explores the AI Ecological Education Policy Framework advanced by Chan, structured into three main dimensions – Governance, Operational, and Pedagogical. The Governance Dimension focuses on ethical governance, while the Operational Dimension examines practical implementation in university settings. The Pedagogical Dimension optimises teaching and learning with AI integration. Qualitative data enriches the chapter, offering a comprehensive perspective for shaping a holistic AI policy in higher education. The chapter also provides a review of AI policies from various global regions, and concludes by offering a step-by-step blueprint for establishing an AI policy within higher educational institutions.

Chapter 4 delves into the transformation of assessment methodologies in higher education, driven by advancements in GenAI. Tracing the evolution of assessment from its traditional roots, the chapter underscores the challenges and innovations brought forth in both pre-GenAI and GenAI contexts. As GenAI offers dynamic, adaptive assessment and feedback mechanisms, the chapter emphasises the paramount importance of maintaining educational principles such as authenticity, integrity, and genuine learning. Introducing the Six Assessment Redesign Pivotal Strategies (SARPS) and the AI Assessment Integration Framework, the chapter highlights nine distinct assessment types tailored for the GenAI era. However, a critical message conveyed is that GenAI should be “co-”, not replace, the human touch in education. The chapter concludes by advocating for a holistic, learner-centric approach to assessment, prioritising feedback and genuine learning experiences over mere grades.

Chan and Colloton’s book is one of the first to provide a comprehensive examination of the use and impact of ChatGPT and Generative AI (GenAI) in higher education. Since November 2022, every conversation in higher education has involved ChatGPT and its impact on all aspects of teaching and learning. The book explores the necessity of AI literacy tailored to professional contexts, assess the strengths and weaknesses of incorporating ChatGPT in curriculum design, and delve into the transformation of assessment methods in the GenAI era. The authors introduce the Six Assessment Redesign Pivotal Strategies (SARPS) and an AI Assessment Integration Framework, encouraging a learner-centric assessment model. The necessity for well-crafted AI educational policies is explored, as well as a blueprint for policy formulation in academic institutions. Technical enthusiasts are catered to with a deep dive into the mechanics behind GenAI, from the history of neural networks to the latest advances and applications of GenAI technologies. With an eye on the future of AI in education, this book will appeal to educators, students and scholars interested in the wider societal implications and the transformative role of GenAI in pedagogy and research.

In this chapter, we peer into the crystal ball of the future, envisioning the profound impact of Generative AI (GenAI) on education. We begin by reflecting on previous technology adoptions and then explore the general predictions surrounding GenAI’s adoption by the population, its potential job impacts, and the safety considerations both pre- and post-adoption. Government and AI development companies enter the spotlight as we contemplate their roles in shaping GenAI’s trajectory. The chapter meticulously examines GenAI’s implications in higher education, spanning pedagogy, research, funding proposals, and journal article submissions. The age-old question of AI’s predictive ability is also probed, with insights from OpenAI ChatGPT, Google Bard, and Microsoft Bing Chat. Throughout this book, we have championed AI’s role in education, highlighting its transformative potential and the need for AI literacy. In this chapter, we take a step further, illuminating the broader societal landscape and its influence on educational institutions and GenAI adoption. As the pages turn, we stand at the precipice of a future where GenAI’s promise and challenges intersect with the world of education.

This chapter offers an in-depth exploration of Generative AI (GenAI) technology within the context of higher education. While the primary focus of this book is on higher education and its implications, we recognise that some readers may have a curiosity about the technical intricacies of GenAI. Therefore, the aim of this chapter is to delve into the technical aspects of GenAI. Beginning with a historical overview of artificial neural networks, it traces their evolution from early developments to recent advances. The chapter unveils the process of creating a GenAI model, covering data gathering, model design, training, and rigorous testing. It also introduces the expansive GenAI ecosystem, highlighting its various applications and interactions within different environments. Furthermore, the chapter spotlights state-of-the-art GenAI models, discussing their applications, design principles, strengths, and limitations. This chapter serves as a foundational knowledge base, equipping readers with essential terminology and understanding of GenAI’s technical aspects.