Learning path · 18 min · beginner
From zero to LLM
The seven concepts that turn "AI" into a real mental model.
You've heard of LLMs. Maybe you've used ChatGPT. This path walks from the foundational ideas (neural networks, embeddings) through the transformer architecture to what an LLM actually is and how it gets trained.
- Artificial IntelligenceAI
Why this step: The umbrella term — to know what AI is, you start by knowing it is not one thing.
Artificial intelligence is the field of computer science that builds systems capable of tasks normally thought to require human intelligence, such as understanding language, recognizing images, or making decisions.
Read full entry →Foundations · beginner - Machine LearningML
Why this step: AI today is mostly machine learning. The shift from rules to learning is the central move.
Machine learning is the branch of AI in which models learn patterns from data instead of being explicitly programmed. The training process adjusts model parameters to reduce error on examples.
Read full entry →Foundations · beginner - Neural Networkartificial neural network
Why this step: The computational building block underlying nearly all modern AI.
A neural network is a stack of simple mathematical units ("neurons") that learn to transform inputs into outputs by adjusting numeric weights during training.
Read full entry →Foundations · beginner - Deep Learning
Why this step: Many-layer neural networks — what made the 2010s breakthrough possible.
Deep learning is machine learning using neural networks with many layers ("deep" = many layers). It powers nearly every recent breakthrough in AI, including LLMs and image generators.
Read full entry →Foundations · beginner - Embeddingvector embedding
Why this step: How a neural network represents meaning numerically. Crucial mental model.
An embedding is a list of numbers (a vector) that represents a piece of input — a word, a sentence, an image — in a space where similar things end up close together.
Read full entry →Architecture · intermediate - Transformer
Why this step: The specific architecture behind every major LLM. Stop here and you understand the dominant 2026 AI.
The transformer is the neural network architecture behind virtually every modern large language model. It uses self-attention to model relationships between all positions in a sequence in parallel.
Read full entry →Architecture · intermediate - Large Language ModelLLM
Why this step: The thing itself. With everything above, the LLM finally makes sense as a layered idea.
A large language model is a neural network trained on huge amounts of text to predict the next token in a sequence. GPT-4, Claude, and Gemini are all LLMs.
Read full entry →Foundations · beginner
You finished the path.
Now stress-test what you remember.