Everything you need to understand generative AI, from what a model is to how it runs in production, in one complete, sequential series. Written to be read by a curious beginner and still useful to an engineer or architect, and vendor-neutral throughout. All 30 parts are now published. Start at Part 1, or jump to the part you need.
Series complete · 30 of 30 parts published
New to AIStart at Part 1 and read in order.
Want to buildJump to Parts 12-22.
Run the infrastructureParts 23-30, the infra track.
Phase 1 · Foundations
- 01What Is Generative AI? A Plain-English Guide
- 02The GenAI Words Everyone Uses, and What They Mean
- 03How We Got from If-Statements to ChatGPT
- 04What a Model Really Is
- 05What Generative AI Can and Cannot Do
Phase 2 · How It Works
- 06How Neural Networks Learn, Without the Math
- 07How Words Become Numbers: Tokens and Embeddings
- 08Attention, the Idea That Made Modern AI Work
- 09Training vs Inference: Why Using AI Is the Real Cost
- 10The Context Window, and Why Models Forget
- 11Why AI Models Make Things Up (and What Temperature Does)
Phase 3 · Using GenAI Well
- 12Prompt Engineering That Actually Works
- 13RAG: How to Stop Your AI Making Things Up
- 14Vector Databases: How Semantic Search Really Works
- 15Fine-Tuning vs RAG vs Prompting: Which One, and When
- 16AI Agents: What Actually Works, and What is Hype
- 17Multimodal AI: Text, Images, and Audio in One Model
- 18Why Looks Good Is Not Enough: Evaluating GenAI Output
Phase 4 · Under the Hood
- 19Why Data, Not Model Size, Usually Decides Quality
- 20Quantization: Running Big Models on Smaller GPUs
- 21Guardrails and Responsible AI: What They Catch, and Miss
- 22Where the Money Actually Goes in Generative AI
Phase 5 · Infrastructure and Serving
- 23Why GenAI Runs on GPUs, and the Memory Wall
- 24vLLM vs TensorRT-LLM vs SGLang: Which Inference Engine
- 25Scaling Inference: Latency vs Throughput (and GPU Ops)
- 26The Network and Storage Behind Large-Scale AI
- 27On-Prem vs Cloud vs Hybrid for GenAI: An Honest Verdict
Phase 6 · Frontier
- 28What It Takes to Train Across Thousands of GPUs
- 29Mixture-of-Experts and Where AI Is Heading
- 30The Economics and Future of Generative AI
Building this on your own infrastructure? See the companion VMware Private AI series for the production build on VCF.
