Interactive reference
The Dependency Map of AI
Understanding the AI family tree allows you to understand AI and how it differs from IT. Reading upward: every layer is built on the ones beneath it, regardless of when it was invented. Foundations sit at the bottom; systems sit at the top. Click any box to trace its full lineage upward to the systems that rely on it, or down to the foundations it rests on. Switch to Compare and pick two architectures to see where their substrate overlaps.
Click any box to trace its full dependency lineage upward to systems, or downward to foundations.
Selected
Compare B
Systems that use it
Systems
0 nodesArchitectures
0 nodesMechanisms
0 nodesPrimitives
0 nodesLearning
0 nodesFoundations
0 nodesHow AI systems behave
AI systems are largely nonlinear in execution, and their results are nearly impossible to predict ahead of time. They rely on small deterministic helper programs so these applications can be managed by deterministic IT systems. It is much easier to understand in context how this works while recently studying linear algebra, probability and statistics and gradient descent. If you have ever looked at how plotting data on graphs works you will develop the intuition to understand how many AI systems work.