AI A to Z: An AI Glossary
- Karen Walstra
- 1 minute ago
- 5 min read
This Artificial Intelligence (AI) glossary from A to Z was created in collaboration with Google’s Gemini to assist in removing the barriers to some of the AI terminology and transform advanced computer science vocabulary into accessible, colourful, and engaging language and visual learning opportunities, regardless of their technical background
A is for Algorithms:
The step-by-step instructions that AI systems follow to learn, solve problems, and make decisions. Think of them as the recipes that guide AI. https://www.coursera.org/articles/ai-algorithms
B is for Bots:
Short for "robots," the automated programs designed to perform specific tasks, often interacting with humans, like customer service chatbots. https://www.kaspersky.com/resource-center/definitions/what-are-bots
C is for ChatGPT:
A famous example of a large language model (LLM) developed by OpenAI, known for its ability to generate human-like text and engage in conversations. https://www.ibm.com/think/topics/chatgpt
D is for Deep Learning:
A powerful subset of Machine Learning (ML) that uses Artificial Neural Networks with multiple layers (deep architectures) to automatically learn complex and abstract patterns from vast amounts of data. It is foundational to modern AI, driving capabilities like Large Language Models (LLMs) and Generative AI - recognising patterns and generating outputs. https://aws.amazon.com/what-is/deep-learning/
E is for Ethics:
The crucial consideration of what's right and wrong when developing and using AI, focusing on fairness, privacy, transparency, and accountability. https://www.coursera.org/articles/ai-ethics
F is for Facial Recognition:
A technology that identifies or verifies a person from a digital image or a video frame by analysing their unique facial features.
G is for Generative AI:
AI systems that can create new, original content, rather than just analysing existing data, such as images, text, music, new content video and even code. https://teaching.pitt.edu/resources/what-is-generative-ai/
H is for Human-in-the-Loop (HITL):
A process where human intelligence is combined with AI to improve the system's performance, often by having humans review and refine AI outputs. Learners, in fact everyone, needs to be the HITL analysing and verifying the AI information. Check, adapt and edit before using or sharing. https://cloud.google.com/discover/human-in-the-loop
I is for Inference:
The process where a trained AI model uses new, unseen data to make predictions, classifications, or generates an output in a real-world setting. It's the AI putting its "learned knowledge" into practice.
J is for Jargon:
Less a technology and more a conceptual challenge. AI has its own jargon (e.g., convolution, backpropagation, stochastic gradient descent). Discussing Jargon in AI is often relevant to transparency and making AI accessible and explainable. The specialised language used in the field of AI, which can sometimes be complex and requires a glossary like this to understand! https://developers.google.com/machine-learning/glossary
K is for Knowledge Graphs:
Structured systems that represent real-world entities and their relationships in a way that AI can easily understand and reason with.
L is for Large Language Model (LLM):
A type of AI model trained on vast amounts of text data to understand, generate, and process human language.
Now, LLMs have moved into Large Multimodal Models (LMMs) or Multimodal Large Language Models (MLLMs): These models are extended to process, understand, and generate information across multiple modalities (types of data), not just text. Accepting image, audio, or video as input and generating relevant text or code, or accepting text and generating an image, e.g., Models like Gemini. https://www.ibm.com/think/topics/multimodal-llm
M is for Machine Learning (ML):
A core component of AI that enables systems to learn from data without being explicitly programmed, allowing them to improve performance over time.
N is for Neural Networks:
Computational models inspired by the human brain's structure, composed of interconnected "neurons" that process information and learn patterns.
O is for OpenAI:
A leading AI research and deployment company known for developing advanced AI models like GPT-4 and DALL-E.
P is for Prompt Engineering:
The art and science of crafting effective inputs (prompts) to guide AI models, especially large language models, to generate desired outputs.
Q is for Quantum AI:
An emerging field that explores how quantum computing principles can be applied to enhance AI capabilities, potentially leading to much more powerful AI.
R is for Robotics:
The branch of engineering and computer science that deals with the design, construction, operation, and application of robots.
S is for Supervised Learning:
A type of machine learning where the AI learns from labeled data, meaning the correct answers are provided during training, helping the AI make accurate predictions.
T is for Training Data:
The dataset used or information to teach an AI model to recognise patterns, make predictions or generate content. The training data consists of examples that the AI learns from to develop its understanding and abilities. https://www.ibm.com/think/topics/training-data
U is for Unsupervised Learning:
A machine learning approach where the AI learns from unlabeled data, identifying patterns and structures on its own, without human supervision and without explicit guidance. https://cloud.google.com/discover/what-is-unsupervised-learning
V is for Vision AI:
AI systems that enable computers to "see" and interpret visual information from images and videos, used in applications like self-driving cars and medical imaging. https://focalx.ai/ai/ai-computer-vision/
W is for Weak AI (Narrow AI):
AI systems designed and trained for a particular, single task, such as playing chess or recommending products. It is goal-orientated and without general human-like intelligence. https://codebots.com/artificial-intelligence/the-3-types-of-ai-is-the-third-even-possible
X is for Explainable AI (XAI):
A set of techniques that make AI models more transparent and understandable, allowing humans to comprehend why an AI made a particular decision.
Y is for "Yawn-free" AI:
The AI developers goal is to create engaging and useful AI tools that don't bore or disengage users, rather keep them interested and productive! https://medium.com/new-writers-welcome/10-neglected-mistakes-you-can-fix-to-create-yawn-free-content-bfea79397021
Z is for Zero-Shot Learning:
An advanced AI capability where a model can perform a task or understand a concept it has never explicitly been trained on, often by leveraging its existing knowledge. https://www.ibm.com/think/topics/zero-shot-learning
🧠 Use the glossary with your learners to:
Build Foundational Vocabulary Fluency:
Understanding core terms is the first step toward true AI literacy.
A glossary provides the fundamental vocabulary required when reading articles, watching videos and to participate meaningfully in conversations about AI.
Improve Conceptual Retention:
Concepts linked to a strong, simple metaphors, such as Training Data depicted as a treasure chest, are easier to commit to long-term memory than abstract technical definitions.
Increase Engagement:
The "fun" elements; the creative analogies and the memorable visuals; makes the learning process more enjoyable.
Empower Independent Learning:
The glossary could be used by learners to quickly reference an unknown term, for example "Zero-Shot Learning", to get an immediate, simple explanation, empowering them to continue their exploration without constantly needing teacher intervention.
Enjoy exploring and discussing with your learners.
I look forward to hearing from you.
Dr. Karen Walstra
All images created by Gemini.
Comments