Skip to content

The AI Family Tree

Run The AI Family Tree MicroSim Fullscreen

About This MicroSim

Seven terms that get used interchangeably in vendor marketing are not interchangeable at all: they nest inside one another like Russian dolls. This MicroSim lays out that nesting as a top-down chain, from Artificial Intelligence at the broadest end down to Large Language Model at the narrowest, with Neural Network attached to the side as the mechanism deep learning is built on rather than as another level of scope.

Node color deepens as scope narrows — pale blue for the broadest category, dark teal for the narrowest — so the nesting is visible before a single label is read. Each term carries the one-sentence definition used in Chapter 1 plus a short note on why that level of the tree matters for cost.

The economic point behind the diagram: "AI" tells you almost nothing about what a system costs to run, because the category stretches from a rule-based script to a model with hundreds of billions of learned values. Every cost estimate in this book attaches to a specific level of this tree.

How to Use

  • Hover any term to read its definition and its cost angle in the right-hand panel.
  • Click a term to highlight the whole chain in orange — every broader category above it, and every narrower category below it. Click it again to clear.
  • Narrow ▶ reveals the tree one level at a time, starting at Artificial Intelligence. Use it to predict what comes next before you see it.
  • ◀ Back steps back up a level; Show All returns to the full tree.
  • Use the on-screen navigation arrows to pan and zoom. Mouse-wheel zoom and drag-panning are deliberately disabled in the chapter embed so the page keeps scrolling normally; both are enabled when you open the MicroSim fullscreen.

Graph Structure

A single "contains" chain runs top to bottom, with one branch off the side.

Nodes

Term Scope Color
Artificial Intelligence Broadest category Palest blue
Machine Learning Subset of Artificial Intelligence Light blue
Deep Learning Subset of Machine Learning Medium blue
Neural Network The mechanism deep learning is built on Lavender (off the scope scale)
Generative AI Subset of Deep Learning Teal
Foundation Model Subset of Generative AI Deep teal
Large Language Model Subset of Foundation Models Darkest teal

Edges

  • contains — solid dark arrows linking each broader category to the next-narrower one it contains: AI → ML → DL → Generative AI → Foundation Model → LLM.
  • built on — one dashed gray arrow from Deep Learning to Neural Network. It is drawn differently on purpose: a neural network is not a narrower kind of deep learning, it is the thing deep learning is made of.

Key Concepts

  • The nesting is strictly one-directional: every LLM is a foundation model, every foundation model is generative AI, every generative AI system is deep learning — and none of those statements reverse.
  • Not every foundation model is an LLM. Some are trained primarily on images, audio, or a mix of data types.
  • The dashed edge marks a category error worth avoiding: "built on" is a different relationship from "is a subset of."
  • Cost lives at specific levels. GPU spend and training time belong to the deep learning level; token pricing and inference cost belong to the LLM level.

Lesson Plan

Grade Level

Undergraduate / professional (business and finance audiences, no technical prerequisites)

Duration

10-15 minutes

Prerequisites

None. This MicroSim is the entry point to the chapter's vocabulary.

Learning Objectives

After using this visualization, students will be able to:

  1. Classify all seven terms according to their nested "is-a-subset-of" relationships.
  2. Explain why "every LLM is a foundation model" is true while the reverse is not.
  3. Distinguish the "built on" relationship (Deep Learning → Neural Network) from the "contains" relationship that links the other six terms.
  4. Identify which level of the tree a given cost claim actually applies to.

Activities

  1. Predict, then reveal (5 min): Start at Artificial Intelligence and use Narrow ▶. Before each click, ask students to name the next-narrower category and defend the guess. Reveal, then compare.
  2. Trace a chain (5 min): Click Large Language Model and read the highlighted chain aloud from bottom to top: "an LLM is a foundation model, which is generative AI, which is deep learning, which is machine learning, which is AI."
  3. Spot the odd edge (3 min): Ask why the arrow to Neural Network is dashed. Students should articulate that it marks a mechanism, not a narrower scope.
  4. Read a vendor claim (5 min): Give students a marketing line such as "our AI-powered platform." Ask which node on this tree the claim actually pins down, and what they would need to ask to pin it down further.

Assessment

  • Can students place all seven terms in the correct nesting order without the diagram?
  • Can students produce a counterexample to "every foundation model is an LLM"?
  • Can students explain why "AI-powered" is not enough information to estimate a running cost?
  • Given a cost statement (for example, "training cost millions in GPU compute"), can students name the level of the tree it belongs to?

Iframe Embed Code

You can add this MicroSim to any web page by adding this to your HTML:

<iframe src="https://Kschuman-aon.github.io/genai-roi/sims/ai-family-tree/main.html"
        height="602px"
        width="100%"
        scrolling="no"></iframe>

Editing Node Positions

To adjust the layout:

  1. Open main.html with the ?enable-save=true parameter.
  2. Drag nodes to the desired positions.
  3. Read the updated coordinates from the network and copy them into the TERMS array in ai-family-tree.js.

References

  1. Chapter 1: Core Concepts of LLMs — the source of every definition in this MicroSim.
  2. Glossary — formal definitions of all seven terms.
  3. vis-network Documentation — the graph library used here.