# How the Tech Industry Works — and What AI Is Changing

This is written at 11:49 at night, so it definitely sounds like techie rubbish in front of you. And I know I didn't write how things are actually going, how this technology really looks, or how the industry really works. I also know there are a lot of grammar and other mistakes.

There are some stories and thoughts in my mind that I just put into this.

Now, the decision is yours: you want to read it or not. It completely depends on you.

And if you read it, judge it. Write in the comments what you think about my thoughts.

Now, the point is:

## How Tech Industry Works

There are basically two kinds of people in this industry.

One is the person who thinks about architecture and develops new things according to the advanced needs of users. This person understands the product deeply. They know about downtime, bottlenecks, scalability, system behaviour, and all the other things that can break when a product becomes bigger.

Then there are implementation-focused developers.

Their speciality is not necessarily product-making or defining the entire architecture. They focus more on implementation, LLD, writing code, and taking ownership of the parts they are responsible for.

That's the basic story.

Now let's come to the next point.

## How AI Came Into This

This is not about what AI is or what AI can do.

AI came into the industry, and people started saying that AI would kill developers.

But I don't think the biggest problem is simply about writing code.

The bigger question is **ownership**.

Some developers are already working on this part, because AI is slowly becoming capable of doing implementation work.

And now AI is also trying to enter the ownership side.

Now you might think AI will not affect the people who think about architecture.

You are wrong.

AI can also be used by technical owners who think, design systems, frame architecture, and try to maintain things properly.

AI can help them think through architecture, compare approaches, find bottlenecks, and maintain systems.

So AI is not only coming for the person who writes the code.

It can also change the way the person who designs the system works.

And now the actual point of this article starts.

## How New Juniors Look Today

I think the result can be garbage.

Yup, no offence.

And don't take this statement too seriously. My thought is going in a different direction here.

But imagine you are Nobita and you have too many gadgets, but you are still dumb.

Yup.

That is the condition of some new joiners today.

They haven't written much code themselves, but they have access to extremely powerful tools.

They can prompt something.

AI gives them the code.

They put it into the project.

It works.

They feel like they have done something amazing.

But now imagine when the real problem comes.

Imagine production goes down at 2 AM.

Now what?

You can't just prompt:

> "Production is down. Fix it."

You need to understand what is happening.

You need to connect the dots.

You need to know where the request is failing, whether the database is the problem, whether the service is overloaded, whether there is a bottleneck, whether something changed in deployment, whether the cache is behaving badly, or whether something completely different is happening.

That understanding doesn't magically come from writing a prompt.

And this is where I have a problem.

YouTube and reels keep showing:

> "Become an architect."

> "Learn system design."

> "Learn LLD."

> "AI can write all the code."

Okay.

But imagine if everyone becomes an architect.

Then who is actually going to be the person building and understanding the implementation?

And if everyone starts thinking that there are 50 different ways to design an LLD for every small problem, things can become even more complicated.

Because architecture is not just about making diagrams.

And coding is not just about writing syntax.

Both need understanding.

## My Experience With AI

For me, this is where things become a little different.

For the last one year, I have used AI more like a conversation partner.

I ask AI things like:

"Which implementation is better?"

"What are the trade-offs?"

"Is this approach actually good?"

But I have not used AI to write my code for me.

In my last API, I contributed around 1000+ lines of code without having AI write the code for me.

And I understand what I wrote.

I am not saying this makes me better than someone who uses AI.

I'm saying I wanted to understand what I was doing before letting AI do it for me.

Maybe I am slower.

Maybe I am wasting some time.

Maybe I could have delivered things faster.

I understand that.

But at least when something breaks, I want my brain to be able to understand the system instead of immediately asking AI what happened.

## Everyone Is Studying AI

Yup.

This is the main point that interests me.

Everyone is studying AI.

Everyone wants AI knowledge.

Everyone wants to use AI.

Everyone wants to become an AI engineer.

Okay.

But if everyone is learning AI, then what is the future?

For me, that might actually be more interesting.

Because if AI knowledge becomes normal, then simply knowing how to use AI will not make you special anymore.

It will become another basic skill.

Then what?

Maybe the difference will be the person who actually understands the system.

Maybe the difference will be the person who knows when AI is wrong.

Maybe the difference will be the person who can take ownership when everything breaks.

Maybe the difference will be the person who understands the product instead of only understanding the prompt.

I don't know.

That's what I am thinking about.

## Some Things I Have Heard About AI

I have heard some interesting comparisons from people.

One person said AI is like an eagle: it is flying at the top, but it is not affordable or accessible to everyone yet.

Another comparison was that AI is like expensive technology in the beginning.

Think about television.

When new technology comes, it is expensive.

Then, day by day, it becomes cheaper and more accessible.

Another person compared AI to the mobile phone.

Mobile phones killed or reduced many old ways of communication, like letter writing.

But at the same time, mobile phones created multiple new things that nobody was thinking about before.

Maybe AI will be the same.

Maybe it will kill some jobs.

Maybe it will create new ones.

Maybe it will remove the value of writing code manually.

Maybe it will increase the value of understanding systems.

Maybe it will change what we call a developer.

Maybe it will change what we call an architect.

Maybe it will make everyone more productive.

Or maybe we are just at the beginning and have no idea what is actually coming.

I don't know.

These are just my thoughts at 11:49 at night.

Now I want to hear yours.

**What do you think?**