Glossary of Artificial Intelligence Terms for Educators

Introduction

• Glossary written by Pati Ruiz and Judi Fusco.
• Designed for educators to reference AI terms.
• Last updated on March 31, 2024.

Key Terms

Artificial Intelligence (AI)

• Branch of computer science.
• Uses hardware, algorithms, and data to create intelligence for decision making, pattern discovery, and actions.
• Built using rule-based systems or machine learning algorithms.

Algorithm

• Core component of AI systems determining decisions.
• Can be rule-based or machine learning algorithms.

Artificial General Intelligence (AGI)

• Not yet realized.
• AI systems that can learn, understand, and solve any human problem.

Artificial Narrow Intelligence (ANI)

• AI for solving narrow problems, e.g., facial recognition.

Generative AI (GenAI)

• A machine learning type that generates content like text, images, music.

ChatGPT Models

• Built on neural network transformer models for natural language processing.
• Generative, Pre-trained, Transformer.

Transformer Models

• Language models used in GenAI.
• Neural networks focusing on important input/output aspects using self-attention mechanisms.

Self-attention Mechanism

• Helps in identifying important features in input.

Large Language Models (LLMs)

• Foundation for GenAI systems.
• Predicts next word based on statistical relationships.

Computer Vision

• Teaching computers to understand visual information.

Critical AI

• Focuses on reflective assessment and critique of AI.

Data

• Information units used by AI.
• Training Data: Used to train algorithms, can perpetuate biases.

Foundation Models

• Large data models as a base for developing AI.
• Controversy over data source trustworthiness.

Human-centered Perspective

• AI systems should augment, not replace, human skills.

Intelligence Augmentation (IA)

• Enhancing tasks, allowing humans to focus on non-redundant tasks.

Intelligent Tutoring Systems (ITS)

• Provide instant feedback to students.
• Supports adaptive learning.

Adaptive Learning

• Adjusts material based on learner's performance.

Interpretable Machine Learning (IML)

• Models that provide their own explanations for decisions.

Black Boxes

• Systems whose internal processes are not visible.

Machine Learning (ML)

• Identifies rules/patterns in data without human intervention.

Neural Networks (NN)

• Inspired by human brain's neuron interconnections.

Deep Learning

• Subset of neural networks with multiple hidden layers.

Natural Language Processing (NLP)

• Helps computers understand and process language.

Robots

• Mechanical machines capable of physical tasks.

User Experience Design/User Interface Design (UX/UI)

• Refers to user interactions with products.

Explainable Machine Learning (XML) or Explainable AI (XAI)

• Helps humans understand ML outputs but not fully trustworthy.

Acknowledgments

• Thanks to Michael Chang, Ph.D., and Eric Nentrup for reviews and support.

Licensing

• Licensed under Creative Commons Attribution 4.0 International License.
• Suggested citation format provided.