Learning hub β€” new lessons weekly

The Late Compiler

It's never too late to learn to think like a machine.

AI, machine learning, Python, and agentic systems β€” explained for people who came to technology later in life. The focus here is applied, not theoretical: how to use these tools, not just understand them. Build something real from lesson one.

5
Learning tracks
~50
Lessons planned
1/week
New content

β€œA compiler translates human language into machine language. Some of us just started the translation a little later than others. That doesn't make the output any less powerful.”

Every lesson here ends with something you can do β€” a tool you can operate, a script you can run, a decision you can make better. The theory comes along for the ride, but the driver's seat belongs to application.

β€” Dr. B.V.R.C. Purushottam

5 learning tracks

Choose your path

Who this is for

You belong here if...

The curious IAS officer

You've heard about AI in governance but never had time to learn the fundamentals.

The 50-year-old professional

Your colleagues use ChatGPT. You want to understand what's actually happening under the hood.

The parent or grandparent

Your children talk about machine learning. You want to talk back β€” intelligently.

The UPSC aspirant

AI is now an essay topic, an interview question, and a policy reality. You need to understand it.

How it works

Learn by doing, not by reading

01

Pick a track

Choose based on what you want to do β€” use AI tools, write code, build models, or understand agents.

02

Do first, understand second

Every lesson starts with a hands-on task. Theory arrives as explanation, not prerequisite. You'll touch real tools from day one.

03

Go at your pace

No deadlines, no exams, no judgment. One article per week. Revisit as often as you need.

04

Build a real portfolio

By the end of each track, you'll have built something β€” a working script, a trained model, or a tool you actually use.

Publishing schedule

What's coming next

Week 1AI fundamentalsWhat is AI, really?
Week 2Python from zeroHello, World β€” your first program
Week 3Applied AIYour first hour with ChatGPT
Week 4Machine learningLinear regression from scratch
Week 5Agentic AIWhat is an AI agent?
Week 6AI fundamentalsThe 5 types of AI
A note from the teacher

I taught myself machine learning at 47. I wrote my first Python script at 47. I built an AI pilot program for 63,000 cooperatives at 48. If I had waited for the β€œright time” to start, none of that would exist.

I learned by doing, not by studying. That's the philosophy of The Late Compiler: every lesson is built around something you'll actually use. You won't sit through hours of theory before touching a real tool. You'll open ChatGPT in lesson one. You'll write working Python by lesson two. The theory arrives when you need it β€” as explanation for something you've already done, not as a prerequisite for something you haven't.

The Late Compiler isn't about catching up. It's about compiling β€” translating the world's most powerful technology into something you can understand, use, and build with. The only prerequisite is curiosity.

See you in class.

Signal from the Frontier
Latest: "The Markov Brain" β€” March 2026

Get the next essay on mind, machine, and meaning

Essays at the intersection of AI, philosophy, and Indian governance. No promotional content.

Views expressed are personal and do not represent the Government of India and Government of Uttarakhand.