Glossary

Adaptive Learning
An instructional method in which software adjusts the path, pace, and level of difficulty of educational materials in real-time based on individual student performance and engagement.

Algorithmic Bias
Systematic errors in AI predictions or recommendations that arise when the data used to train a model do not represent all student populations fairly. In K–12, algorithmic bias can affect grading, student grouping, or access to resources.

Artificial Intelligence (AI)
The field of computer science that enables machines to mimic human intelligence processes such as learning, reasoning, and problem-solving.

AI-Driven Proctoring
Technology that monitors students during online or remote assessments using AI to detect possible cheating or unusual behaviors (e.g., looking away frequently, multiple people in view). While it ensures academic integrity, it also raises important questions about privacy and the student experience.

Assistant
In the context of AI in education, an "assistant" refers to an AI-powered tool or system designed to support teachers, students, and administrators in various educational tasks. These assistants can take different forms, including chatbots, virtual tutors, automated grading systems, and personalized learning platforms.

Attention
The attention mechanism in AI language models, especially within transformer architectures, is a process that dynamically assigns weights to different parts of the input data, allowing the model to focus on the most relevant tokens when processing information. Instead of treating every word equally, attention computes a weighted representation of the entire sequence so that, for each token, the model can identify and emphasize other tokens that provide crucial contextual clues.

Automated Summarization
The use of natural language processing (NLP) techniques to condense lengthy or complex documents into concise summaries. For students, automated summarization can produce quick overviews of reading materials, while teachers can use it to generate study guides or check comprehension of lesson content.

Augmented Reality (AR)
An interactive experience that overlays digital elements—images, sounds, or text—onto the real world. AI-driven AR can dynamically adjust these elements for each learner’s needs, creating enriched, hands-on learning scenarios.

Automated Grading
The use of AI algorithms to evaluate student work—especially quizzes, essays, or short answers—saving teacher time while offering instant feedback. Care must be taken to ensure fairness and accuracy.

Bias (in AI)
Systematic error or prejudice in the output of an AI model, often reflecting biased training data or flawed design assumptions. AI bias can lead to unfair or discriminatory outcomes and is a significant concern in real-world applications

Chatbot
An AI program capable of engaging in text or voice-based interactions. In K–12 settings, chatbots can answer routine questions, give quick feedback, or assist with administrative tasks (e.g., assignment deadlines).

Cloud Computing
The on-demand availability of computing services, such as data storage and processing power, over the internet. Schools can leverage cloud-based AI services for resource-intensive tasks without needing local servers.

Competency-Based Education
An educational model that measures learning by demonstrated mastery of specific skills or competencies rather than seat time or traditional grade levels. AI can support this by tracking individual progress in real time and offering just-in-time activities to help students achieve clearly defined performance benchmarks.

Computer Vision
An AI domain where algorithms interpret and understand visual information from the world, such as images and videos. In education, it can support applications like automated grading of handwritten work or real-time object recognition for interactive lessons.

Context Window
A "context window" in AI refers to the limited amount of information an AI model can process at one time, essentially acting like a "short-term memory" where the model considers surrounding text or data points to understand the context of a query or prompt, allowing it to generate more relevant and coherent responses; the larger the context window, the more information the AI can process simultaneously, leading to better comprehension of complex topics or longer conversations

Data-Driven Instruction
An approach where teaching decisions (e.g., lesson planning, interventions) are guided by analysis of student data, which might include performance trends flagged by AI tools.

Deep Learning
A specialized branch of machine learning using algorithms called neural networks, which mimic the human brain’s structure to recognize patterns in large datasets—helpful for tasks like image recognition or language translation.

Digital Content Curation
The process of automatically gathering, organizing, and recommending digital resources aligned to curricula or individual student profiles. AI-driven content curation helps educators and students access high-quality, relevant materials more efficiently, supporting a more personalized and effective learning experience.

Digital Divide
The gap between students who have easy access to modern information and communication technology (including AI-powered tools) and those who do not. It highlights disparities in device availability and internet connectivity.

Distillation
LLM distillation is a pivotal technique in making large language models more practical and efficient. By transferring essential knowledge from a complex teacher model to a smaller student model, distillation preserves performance while reducing size and computational demands. Deepseek R1 uses distillation to require far few resources than Chat GPT.

Educational Data Mining (EDM)
A field related to learning analytics that applies data mining techniques (like clustering, classification, or regression) to educational data. It focuses on discovering patterns to improve teaching and learning experiences.

Ethical AI
The practice of designing and deploying AI systems that prioritize fairness, transparency, and privacy. In schools, Ethical AI ensures student data is protected and decisions do not discriminate or stereotype.

Freemium
The freemium pricing structure is a business model in which a software or app is offered for free with limited features, while additional premium features, enhancements, or content are available for purchase. This model allows users to access the basic functionality at no cost, encouraging widespread adoption, while generating revenue from users who choose to upgrade for advanced capabilities, ad-free experiences, or exclusive content.

Gamification
The incorporation of game-like elements—points, badges, leaderboards—into educational experiences. AI can personalize these game elements (e.g., difficulty levels) for each learner.

Generative AI
A category of AI models that create new content—such as text, images, audio, or video—instead of merely analyzing or classifying existing data. Popular examples include text generation models (like ChatGPT) and image synthesis models (like DALL·E).

Generative Fill
Generative Fill in AI refers to an AI-powered feature that allows users to modify, expand, or remove parts of an image using generative models. It analyzes the surrounding context of an image and seamlessly fills in missing or altered areas with realistic, AI-generated content.

Hallucination
In the context of artificial intelligence (AI), hallucination refers to instances where an AI system, particularly large language models (LLMs) like ChatGPT, generates outputs that are incorrect, misleading, or entirely fabricated, despite appearing plausible and coherent. These inaccuracies can manifest as factual errors, nonsensical statements, or the introduction of information not grounded in the model's training data.

Intelligent Tutoring System (ITS)
An AI-driven software application designed to provide personalized instruction and feedback, replicating one-on-one tutoring experiences for students in various subjects.

ISTE
The International Society for Technology in Education (ISTE) is a nonprofit organization that helps educators use technology to revolutionize learning.

Large Language Model (LLM)
A type of transformer-based model with a large number of parameters, trained on vast amounts of text data. LLMs are capable of generating coherent text, answering questions, translating languages, and more, often with minimal task-specific training.

Learning Analytics
The collection and analysis of student performance and engagement data to inform teaching strategies. AI-powered analytics can reveal patterns (e.g., frequent misconceptions) and help educators adjust instruction accordingly.

Learning Management System (LMS)
A platform that organizes and delivers educational content, tracks student progress, and manages resources. When enhanced with AI, an LMS can provide personalized learning paths, automated grading, and real-time insights into student engagement.

Machine Learning (ML)
A subset of AI focused on creating algorithms that allow computers to learn from data and improve their performance over time without being explicitly programmed.

Natural Language Processing (NLP)
A branch of AI that deals with the interaction between computers and human (natural) languages. In K–12, NLP can power tools like chatbots, grammar-checkers, or text analysis platforms.

Neural Network
A computing system inspired by the human brain’s interconnected cells (neurons). Neural networks can learn from examples to make predictions or classifications (e.g., recognizing handwritten letters).

Personalized Learning
A broader teaching approach that tailors instruction, content, and pace to each student’s needs, often supported by AI tools that track student data and recommend customized activities.

Predictive Analytics
AI-driven methods for analyzing current and historical data to make predictions about future outcomes. In K–12, this might be used for early identification of students at risk of struggling academically.

Pre-trained Transformer
A transformer model that has been trained on a massive amount of data (e.g., large text corpora) to learn general language representations before being fine-tuned for specific tasks. Examples include BERT, GPT, and RoBERTa.

Prompt

In the context of AI in schools, a prompt is the input or question given to an AI system to guide its response. A well-crafted prompt helps students and educators interact effectively with AI tools to generate useful answers, explanations, or creative content. For example, a teacher might use a prompt like "Explain the water cycle in simple terms for a 4th-grade student."

Prompt Engineering

Prompt engineering is the practice of designing and refining prompts to optimize AI-generated responses. In schools, this skill helps students and teachers get more accurate, relevant, and insightful answers from AI tools. It involves structuring questions clearly, adding context, or specifying response formats. For instance, instead of asking "Tell me about planets," a student might refine it to "List the planets in our solar system and describe one key feature of each."

Privacy & Data Protection
Safeguards to ensure that students’ personal and academic information is collected, stored, and used responsibly. AI applications in education must comply with regulations (e.g., FERPA in the U.S.) to protect student data.

Sentiment Analysis
An AI technique, often part of natural language processing, used to detect and interpret emotions in text. In classrooms, it can help educators gauge student sentiment in discussions or essays, highlighting areas where additional support may be needed.

Self-Regulated Learning with AI Support
The practice of using AI tools and analytics to help students plan, monitor, and evaluate their own learning. These tools might prompt students to set goals, track their progress against personalized targets, and reflect on study habits, fostering independence and critical thinking skills.

Speech Recognition
A type of AI technology that identifies spoken language and converts it into text. In K–12, it can assist students with dyslexia, support language learning, and help teachers quickly transcribe lessons.

Text-to-Speech (TTS)
Technology that converts written text into spoken audio. In K–12, TTS can support learners with reading difficulties or those studying new languages by providing an auditory representation of written materials.

Token
In Natural Language Processing (NLP), a token is the smallest individual unit of text that a model processes. Depending on the application, tokens can be words, subwords, or characters. For example, the sentence “Hello world!” might be split into tokens [“Hello”, “world”, “!”]. The number of allowable tokens affects how large a body of text (the context window)a model can handle.In the original GPT-4o, users could input up to 124,000 tokens and receive up to 4000 tokens in response. With GPT-4o long output, users can now input up to 64,000 tokens and potentially receive up to 64,000 tokens in return.Aug 7, 2024

Virtual Laboratory
A simulated environment where students can perform experiments and practice skills. AI enhancements adapt lab tasks in real time, offering personalized hints or corrective feedback to deepen understanding of scientific concepts.

Virtual Reality (VR)
An immersive technology that simulates a three-dimensional environment. AI can enhance VR experiences by personalizing or adapting digital content based on real-time student feedback or interactions.

Virtual Tutor
An AI-enabled tutoring platform that simulates a human tutor’s guidance. It provides feedback, answers questions, and suggests learning resources independently of teacher supervision.

Voice Assistant
An AI-powered system that interprets and responds to spoken commands. In K–12 settings, voice assistants can help students with quick research, language practice, or controlling classroom devices, promoting hands-free interaction.