AI Glossary and References
Reference list and diagrams of common terms related to AI (artificial intelligence) and ML (machine learning), with two key takeaways about why AI and ML matter to us all.
AI usually stands for “Artificial Intelligence”. The term often gets used in inconsistent ways, though. The goal of this Glossary is to serve as a quick reference to help folks make sense of what they read and hear about “AI”.
Audience for this post: Anyone. Zero prior knowledge of AI is required.
Types of Artificial Intelligence
Within the world of AI, we hear a lot of acronyms and terms. Here are some of the most common terms you’ll hear.
AGI — Artificial General Intelligence (sometimes called “strong AI”)
AI — usually “Artificial Intelligence” (but it’s an overloaded term, and sometimes it gets used to describe products that don’t really have “AI inside”)
[A]NN — [Artificial] Neural Networks
DL — Deep Learning (e.g. a Deep Neural Network with > 3 layers)
GenAI — Generative AI (creates new content)
ML — Machine Learning
NLP — Natural Language Processing (sometimes, but not always, involving AI)
RL[HF] — Reinforcement Learning [with Human Feedback]
Figure 1 shows how these terms relate to each other. We’ll briefly cover them all in AI Fundamentals article #1.
🗝️ KEY TAKEAWAY: Generative AI is only a small subset of everything that can be considered “AI”. When you read announcements or criticisms of “AI”, be sure to ask whether they are talking about AGI, just the GenAI subset, or other types of AI. It makes a difference in terms of capabilities, risks, and ethics.
Types of AI Models
Texts in white in Figure 1 are specific kinds of AI models, including ML and GenAI. If you’re not a data scientist or a software developer working on building an AI-based tool, don’t worry too much about understanding the details and differences. AI Fundamentals #2 will cover what you need to know as a potential user of AI tools based on these models. Here’s what the terms mean:
CNN — Convolutional Neural Network (special type of NN model; commonly used for OCR (Optical Character Recognition) and other purposes)
Diffusion models (used for images and music - examples: DALL-E, Riffusion)
GAN — Generative Adversarial Network (set of special NNs used for genAI, such as for images and video)
GPT — Generative Pre-trained Transformer (examples: ChatGPT, GPT-4, Copilot, BERT, Bard, Midjourney)
LLM — Large Language Model (a form of generative AI, often using a GPT, that is trained on a huge body of texts and has billions of parameters)
RAG — Retrieval-Augmented Generation (an information retrieval method sometimes used together with a LLM to focus on a specific set of source data)
RNN — Recurrent Neural Network (type of NN ML that is often used for Text To Speech, or TTS)
VAE — Variational AutoEncoder (often used for image recognition, natural language processing, and anomaly detection)
Figure 2 shows how all of the different “domains” where generative AI can be used. Text, images, speech, music, and video are all possible. So are other applications you might not think of, including writing software source code or generating synthetic data for training an AI-based tool.
For more on use of generative AI for music, see this series of articles on ethics of genAI for music.
For more on generating synthetic data, see this article by .
Bottom Line
The most important things to know about AI and ML are:
AI and ML are already everywhere in our daily lives. Not convinced? See: “But I Don’t Use AI": 8 Sets of Examples of Everyday AI, Everywhere
AI and ML run on data - yours, mine, everyone’s. Knowing where and how an AI / ML system got its data is important. The humans who created the data an AI system uses are morally and ethically entitled to “consent, control, credit, and compensation” for its use. See: “Giving credit for the 3Cs / 4Cs where due: an origin story” (coming soon)
What’s Next?
We hope that this article serves as a useful reference. If you’re curious to learn more about AI, please see these AI Fundamentals posts:
What questions do you have about AI that we can answer next?
Really appreciate the diagrams too!