FIELD NOTES - GENERATIVE AI
Generative AI,
explained end to end.
What it is, why it matters, how the math works, how to build one yourself, and where the field is headed - all in one page, written like a proper notebook rather than a marketing brochure.
What is GenAI, why use it, why it's needed
What is Generative AI?
Generative AI is a class of machine learning models that create new, original content — text, images, audio, video, or code — instead of only labeling or predicting a category for existing input. It learns the statistical patterns of huge datasets well enough to produce brand-new samples that follow those same patterns.
Why use it?
It compresses hours of writing, coding, designing, or researching into seconds. A single model can draft an email, debug a function, design an illustration, or summarize a 300-page report — all through plain natural language.
Why is it needed?
Data and tasks now scale far beyond what humans can process manually. GenAI gives individuals and small teams the kind of leverage that used to require entire departments — answering support tickets at scale, scanning research literature, or exploring design variations instantly.
Discriminative vs Generative
A discriminative model answers "is this spam or not?" A generative model answers "write me an email like this" — it models the full data distribution P(x), not just a decision boundary between classes.
Take these notes with you
Every section on this page, bundled into a single Markdown file.