The building blocks — start here to get a confident mental model of how it all works.
The foundational mechanics of modern AI models
Transformers · Attention Mechanisms · Tokens · Embeddings
The dominant architecture behind modern AI, parallelizable, scalable, attention-powered
How models decide what to focus on, the core of what makes transformers powerful
The atomic units of language that AI models actually process
How models represent meaning internally, dense vectors that capture semantic relationships
How LLMs are built, from raw data to capable, safe assistants
Pre-training · RLHF / Post-training · Reasoning Training / RLVR
The massive first stage that teaches a model to predict language
Turning a raw autocomplete engine into a useful, safe assistant
How modern models learn to think step-by-step using verifiable rewards
The three families of LLMs and what distinguishes them
Base Models · Instruction-tuned Models · Reasoning Models
Pure autocomplete, powerful but raw and hard to direct
The chatbots you know, RLHF-trained to follow instructions reliably
The current frontier, models that think before they answer
The vocabulary every AI practitioner needs to operate confidently
System Prompts · Context Windows · Parameters · Training vs. Inference · Hallucinations · Jailbreaking
The instructions that shape how a model behaves before a user says anything
The maximum text a model can see at once, and why it's so hard to extend
The learned weights that define a model, and why size matters
Building a model vs. running one, fundamentally different compute profiles
When LLMs confidently state things that aren't true, and why it's a fundamental problem
Bypassing a model's safety training through adversarial prompting
What modern LLMs can actually do beyond generating text
Tool Use · Agentic Capabilities · Multimodality
How LLMs interact with external systems to extend their capabilities
LLMs running in loops with tools to complete multi-step tasks autonomously
AI that can see, hear, and reason across text, images, audio, and more
The competitive landscape, business models, and vocabulary of the AI industry
Major AI Players · Open Source vs. Open Weights · What a Wrapper Is
Who's building frontier AI and what differentiates each lab
The important distinction between truly open AI and 'open enough'
The thin AI product problem, and why it matters for business strategy
Hands-on techniques for working with AI systems
Prompt Engineering · API Basics · RAG (Retrieval-Augmented Generation)
The craft of writing inputs that reliably get the outputs you want
How to actually call an LLM from code, the mechanics every builder needs
Grounding AI responses in your data, the go-to pattern for custom knowledge bases