Agentic AI, As I Understand It

Act 0: Intro (Or How I Ended Up Here, Explaining Stuff I Thought I Knew)

Jul 02, 2025

Hello dear reader,

Before we dive in, let’s get one thing straight:
I’m not a self-declared “AI thought leader”, “AI Expert”, or “AI Influencer”.

However, I am someone who has been working in AI for over five years.
I hold a Master’s degree in the field, and I’ve completed projects, internships, and endured late nights of debugging and wondering if the bug is in my code… or in me.

Still, even with all that, an internal voice inside me constantly repeats:

“Wait, are you even understanding what you’re doing?”

Why Am I Writing This?

What I have is:

A laptop
A brain (which sometimes works)
And a deep conviction:

The best way to learn something deeply… is to teach it.
And the second-best way?
Write about it publicly so people can correct you when you’re wrong. 💀

That’s what this series is about:
A way for me to refine what I’m learning, especially as I go through an intense AI agent engineering bootcamp.
(Shoutout to Meri and Hai for encouraging and inspiring me to finally launch this!)

Why “As I Understand It”?

Because I’m not here to pretend I have all the answers, but to ask the right questions.
I’m here to think out loud, to break down what I think I understand, write about it, and explain it as clearly as I can.

“As I Understand It” is an honest motto.
I’ll use it in every title to reflect what a normal learning journey looks like: From hype to doubt, from misconception to understanding, and back again.

Who Is This For?

If you’re just getting started in AI, welcome.
If you’re already working in the field but secretly googling terms after meetings (like I often do), welcome too.
And if you just want to read a technical blog that doesn’t sound like a PhD thesis and doesn’t treat you like an idiot?

You’re in the right place.

But Wait… What Is an Agent?

Good question.
At a glance?
It’s not James Bond.
It’s what happens when you delegate a task to a language model, equip it with some tools, and give it the terrifying power to make decisions independently.

Think of it like this:
ChatGPT drinks three Red Bulls, picks up a to-do list, and confidently says:

“Don’t worry, boss. I’ll handle it.”

That’s an agent.
It plans, it acts, it reacts. Sometimes brilliantly…
And sometimes, like my cousin, who once broke the central unit of his computer by inserting a coin into the CD reader… to pay for an online game.

We’ll get into all that. Eventually. In the following articles of this series. Stay tuned!

Final Word

Dear reader,
This isn’t a lecture.
It’s a journey, mine and yours, from “I have no idea” to “huh, maybe I get it now.”

And you’re invited.

So welcome to Agentic AI, as I Understand It.
Let’s make confusion a feature, not a bug!