Foundations
The atomic units of intelligence — master the core concepts and mathematical pillars that sustain modern AI.
Foundations: The Atomic Units of Intelligence
Before progressing through the seven levels of AI fluency, you need to master the underlying mechanics of large language models and the mathematical foundations that make them possible. The shift from traditional software to AI represents a transition from deterministic logic to probabilistic prediction — and understanding this shift is the key to everything that follows.
What You Will Learn
The Foundations modules are divided into two tracks:
Core Concepts
These three modules cover the essential mechanics of how language models work. You do not need any mathematical background to complete them.
Mathematical Foundations
These four modules cover the mathematics that sustain modern AI. They are designed to build intuition through practical examples rather than formal proofs. You can complete them in any order, though the sequence below is recommended.
Suggested Approach
-
Start with Core Concepts (F.1 through F.3). These give you the vocabulary and mental models for understanding everything else in the curriculum.
-
Then work through the Math Foundations (F.4 through F.7). Even if you find the math challenging, focus on building intuition. You do not need to master every formula — understanding what the math does and why it matters is more valuable than symbolic manipulation.
-
Revisit as needed. These modules are designed to be references you return to. As you progress through higher levels, you will find that concepts from the Foundations take on deeper meaning.
Why Foundations Matter
The seven levels of AI fluency build on these foundational concepts like a tree grows from its roots. Without understanding tokens, you cannot reason about prompt costs or multilingual performance. Without understanding context windows, you cannot design effective RAG pipelines. Without understanding probability, you cannot interpret model outputs with appropriate confidence. And without the mathematical foundations, you cannot understand — or critically evaluate — the architectures and algorithms that power modern AI.
The time you invest here pays compound interest throughout the rest of the curriculum.