Learn Modern AI
A high-contrast guide and lightweight decision game for learning the systems behind modern AI: transformers, RAG, MoE, diffusion, agents, reasoning, and physical AI.
Understanding Modern AI
Artificial Intelligence has moved at a breakneck speed since the introduction of the Transformer architecture. This site explains the key engineering ideas behind modern models without the math-heavy clutter, updated for the 2026 shift toward reasoning models, agentic workflows, long-context systems, efficient open-weight models, and physical AI.
Read the chapters in order or play AI Systems Quest, a compact scenario game where each choice tests whether you know when to use retrieval, attention, sparse experts, tool calls, or extra test-time compute. Both paths are optimized for Kindle and other e-ink displays.
The Transformer Core
Understand self-attention, the Query-Key-Value mechanism, and how it replaced recurrent networks to form the bedrock of generative AI.
LLM Training & Alignment
Deep dive into pre-training, fine-tuning, and alignment mechanisms like RLHF and DPO that make AI models helpful and controllable.
RAG & Context Windows
Explore how models fetch real-time data using Vector Databases, semantic search, and the mechanics behind ultra-long context windows.
Scaling Efficiency: MoE & Quantization
How Mixture of Experts (MoE) keeps models fast by only activating subset networks, and how quantization compresses parameters for consumer hardware.
Diffusion & Generative Media
Learn how diffusion models generate images and videos by systematically removing random noise, and how Latent Diffusion speeds this up.
Agentic AI & Reasoning
Step into loop-based reasoning, tool usage (function calling), and System 2 thinking paradigms where models reason before responding.
Future Frontiers & Physical AI
Analyze native multimodality, synthetic data constraints, and the next physical frontiers for AI integration.