What is AI ? Building Your First Model in Python Hands-on Lab
AI is software that learns from examples, not fixed rules — like a child learning "dog" after seeing 50 dogs. Every AI follows the same 3 simple steps: Data (experience), Model (brain), Learning (practice with model.fit()). You build your first AI with 2 files: train.py saves model.pkl, predict.py loads it — train once, predict 1000 times.
A Morning in Solan
It's 6:30 AM. The mist is still sitting on the hills of Solan, the chai tapri is just starting to boil its first kettle, and your phone lights up.
You pick it up, and it unlocks the moment it sees your face - even before your first sip of chai. You mumble, "Play some old Kishore Kumar," and it gets it right. On your way to get bread, Google Maps says, "Traffic is light on Mall Road, you'll reach in 7 minutes."
You didn't program any of that. You didn't tell your phone how your half-sleepy face looks, or how to understand your voice.
So who did? That's where our story begins.
What AI Really Is
Forget what ever you have seen in the the sci-fi movies for a minute.
Artificial Intelligence is just software that learns from examples to make a decision, instead of following only fixed rules we wrote.
Traditional software is like a recipe: If A happens, do B. Always.
AI software is different: I showed you 10,000 examples of A and B, now you figure out the pattern yourself. And it does.
Code Lab: Train and Predict
Code Lab: Train and Predict Locked
Sign in with Google to read this architectural note.
chapter-01-what-is-ai.zip
11.6 KB • Pro Member Vault