Appendix G — Glossary

Terms are listed alphabetically.

G.1 A

AI (Artificial Intelligence): Software systems that can perform tasks that typically require human intelligence, such as pattern recognition, decision-making, language understanding, and problem-solving.

AI Ethics: The study of moral principles and guidelines for the responsible development and use of artificial intelligence systems.

AI Literacy: The ability to understand, evaluate, and effectively use AI systems and their outputs.

Assessment Rubrics: Structured evaluation criteria used to assess student work against specific learning outcomes and standards.

G.2 B

Bias in AI: Systematic errors in AI systems that result in unfair or discriminatory outcomes, often due to biased training data or algorithmic design.

G.3 C

Chain of Thought: Generated intermediate reasoning text. A plausible explanation is not a guarantee of a correct answer or a faithful account of the model’s internal computation. This book’s stepwise tutoring asks the learner to attempt and justify steps.

CRAFT Framework: A structured approach to writing effective prompts, consisting of: - C: Context (background information) - R: Role (AI persona to adopt) - A: Action (specific task to perform) - F: Format (desired output structure) - T: Tone/Target (intended audience and style)

Critical Engagement: The practice of actively questioning, evaluating, and critiquing AI outputs rather than accepting them passively.

G.4 D

Deep Learning: A subset of machine learning using neural networks with multiple layers to process complex patterns and data.

Dialogue-Based Assessment: Assessment methods that evaluate student thinking through interactive conversations rather than static products.

G.5 E

Evidence-Based Practice: Professional decision-making grounded in research, data, and systematic evaluation rather than intuition alone.

G.6 F

Few-Shot Prompting: Supplying a few examples in a prompt to guide the response. This does not necessarily update the model’s trained parameters.

G.7 G

Generative AI: AI systems that can create new content, such as text, images, or code, rather than just analysing existing data.

G.8 H

Hallucination: When AI generates false or misleading information confidently, as if it were factual.

Human Oversight: The practice of humans reviewing, validating, and intervening in AI processes to ensure accuracy and ethical outcomes.

G.9 I

Iterative Refinement: The process of repeatedly improving AI outputs through feedback and revision cycles.

G.10 L

Large Language Models (LLMs): Models trained on language data to generate and process text. A model is distinct from the product that supplies its interface, retrieval, memory and account controls.

Learning Outcomes: Specific statements describing what students should know, understand, or be able to do after completing a learning experience.

G.11 M

Machine Learning: A type of AI where systems learn patterns from data and improve performance without being explicitly programmed for each task.

Meta-Prompting: Using AI to help you create better prompts for AI, essentially using AI to improve your AI interactions.

Metacognition: Awareness and control of one’s own learning processes, including planning, monitoring, and evaluating learning strategies.

G.12 P

Process-Based Assessment: Evaluation methods that focus on how students think and work through problems, rather than just the final product or answer.

Product-Based Assessment: Traditional evaluation methods that focus primarily on the final output or result, rather than the thinking process.

Prompt Engineering: The practice of crafting effective instructions (prompts) to get desired outputs from AI systems.

Prompting: The act of providing instructions or questions to AI systems to elicit specific responses or behaviours.

G.13 R

Retrieval-Augmented Generation (RAG): Providing retrieved material as context for generation. The material need not be current, accurate or complete; the response still requires verification.

Rubrics: Detailed scoring guides that specify criteria for different levels of performance on assessment tasks.

G.14 S

Stepwise Tutoring: A practice pattern in which a learner attempts one step, receives checked feedback and explains a correction before continuing.

Scaffolding: Educational support structures that help students achieve tasks they couldn’t accomplish independently, gradually removed as competence develops.

Self-Assessment: The process where students evaluate their own work and learning progress against established criteria.

G.15 T

Transparency Model: An approach that asks learners to disclose relevant permitted assistance and explain consequential decisions. It need not require complete private chat histories and does not by itself establish authorship.

Transfer Learning: AI’s ability to apply knowledge learned from one task to perform well on related tasks.

G.16 V

Virtual Company: Simulated business environments created through AI conversations, allowing students to practise bounded professional scenarios. Simulation still requires safeguards and debriefing.

# Glossary Terms are listed alphabetically. ## A **AI (Artificial Intelligence)**: Software systems that can perform tasks that typically require human intelligence, such as pattern recognition, decision-making, language understanding, and problem-solving. **AI Ethics**: The study of moral principles and guidelines for the responsible development and use of artificial intelligence systems. **AI Literacy**: The ability to understand, evaluate, and effectively use AI systems and their outputs. **Assessment Rubrics**: Structured evaluation criteria used to assess student work against specific learning outcomes and standards. ## B **Bias in AI**: Systematic errors in AI systems that result in unfair or discriminatory outcomes, often due to biased training data or algorithmic design. ## C **Chain of Thought**: Generated intermediate reasoning text. A plausible explanation is not a guarantee of a correct answer or a faithful account of the model's internal computation. This book's stepwise tutoring asks the learner to attempt and justify steps. **CRAFT Framework**: A structured approach to writing effective prompts, consisting of: - **C**: Context (background information) - **R**: Role (AI persona to adopt) - **A**: Action (specific task to perform) - **F**: Format (desired output structure) - **T**: Tone/Target (intended audience and style) **Critical Engagement**: The practice of actively questioning, evaluating, and critiquing AI outputs rather than accepting them passively. ## D **Deep Learning**: A subset of machine learning using neural networks with multiple layers to process complex patterns and data. **Dialogue-Based Assessment**: Assessment methods that evaluate student thinking through interactive conversations rather than static products. ## E **Evidence-Based Practice**: Professional decision-making grounded in research, data, and systematic evaluation rather than intuition alone. ## F **Few-Shot Prompting**: Supplying a few examples in a prompt to guide the response. This does not necessarily update the model's trained parameters. ## G **Generative AI**: AI systems that can create new content, such as text, images, or code, rather than just analysing existing data. ## H **Hallucination**: When AI generates false or misleading information confidently, as if it were factual. **Human Oversight**: The practice of humans reviewing, validating, and intervening in AI processes to ensure accuracy and ethical outcomes. ## I **Iterative Refinement**: The process of repeatedly improving AI outputs through feedback and revision cycles. ## L **Large Language Models (LLMs)**: Models trained on language data to generate and process text. A model is distinct from the product that supplies its interface, retrieval, memory and account controls. **Learning Outcomes**: Specific statements describing what students should know, understand, or be able to do after completing a learning experience. ## M **Machine Learning**: A type of AI where systems learn patterns from data and improve performance without being explicitly programmed for each task. **Meta-Prompting**: Using AI to help you create better prompts for AI, essentially using AI to improve your AI interactions. **Metacognition**: Awareness and control of one's own learning processes, including planning, monitoring, and evaluating learning strategies. ## P **Process-Based Assessment**: Evaluation methods that focus on how students think and work through problems, rather than just the final product or answer. **Product-Based Assessment**: Traditional evaluation methods that focus primarily on the final output or result, rather than the thinking process. **Prompt Engineering**: The practice of crafting effective instructions (prompts) to get desired outputs from AI systems. **Prompting**: The act of providing instructions or questions to AI systems to elicit specific responses or behaviours. ## R **Retrieval-Augmented Generation (RAG)**: Providing retrieved material as context for generation. The material need not be current, accurate or complete; the response still requires verification. **Rubrics**: Detailed scoring guides that specify criteria for different levels of performance on assessment tasks. ## S **Stepwise Tutoring**: A practice pattern in which a learner attempts one step, receives checked feedback and explains a correction before continuing. **Scaffolding**: Educational support structures that help students achieve tasks they couldn't accomplish independently, gradually removed as competence develops. **Self-Assessment**: The process where students evaluate their own work and learning progress against established criteria. ## T **Transparency Model**: An approach that asks learners to disclose relevant permitted assistance and explain consequential decisions. It need not require complete private chat histories and does not by itself establish authorship. **Transfer Learning**: AI's ability to apply knowledge learned from one task to perform well on related tasks. ## V **Virtual Company**: Simulated business environments created through AI conversations, allowing students to practise bounded professional scenarios. Simulation still requires safeguards and debriefing.