AI Made Me 50x Faster — And Somehow More Lazy

How I stopped outsourcing my brain and started thinking with AI instead.

When I first found AI three years ago, I thought I’d found a cheat code for life.

I used it for everything. Writing emails. Drafting content. Planning projects. Researching topics.

The first few months felt incredible. Tasks that used to take me hours were done in minutes. I’d paste a problem into ChatGPT, get a solution, implement it, and move on to the next thing.

I felt like I’d unlocked some secret superpower.

But around month three, something strange started happening.

I’d sit down to write an email explaining a project update and immediately think:

“Why struggle with this? Just let ChatGPT draft it.”

What happened was I reached for AI before I even tried to understand the whole project context. I felt this pull to outsource the thinking before I’d done any thinking myself.

I wasn’t using AI to amplify my capabilities anymore. Instead, I was using it to avoid developing them.

The worst part? I didn’t notice it happening.

I used to be able to sit with a hard problem for hours, turning it over in my mind, exploring different angles, building understanding through struggle.

Now, the moment something got difficult, my brain immediately suggested:

“Just ask ChatGPT. It’ll figure it out in 30 seconds.”

And it would. That was the problem.

Because every time I took that shortcut, I lost a little bit of the muscle that made me good at this in the first place.

I realized I’d fallen into the trap that nobody talks about when they’re selling you on AI productivity: I was outsourcing tasks when I should have been thinking in tandem with AI.

There’s a massive difference between these two approaches:

• Outsourcing to AI: “Do this for me” → Get result → Move on → Lose capability over time

• Thinking with AI: “Help me understand this” → Learn the framework → Build transferable skills → Get stronger over time

I had been doing the first thing exclusively. And it was making me softer, not sharper.

So I rebuilt my entire approach. I stopped using AI as a shortcut and started using it as a thinking partner. I stopped asking it to do things for me and started asking it to help me understand how to think about problems.

I Hit an AI Ceiling I Didn't Know Existed (Here's How I Broke Through)

That shift changed everything—not just how I use AI, but how I approach learning, problem-solving, and skill development.

Today’s guest post is from Orel, who builds WriteStack: an AI tool for Substack creators to grow using Notes. He’s spent the past year discovering the same pattern I did, both in his own work and watching hundreds of creators fall into the same trap.

This post isn’t about abandoning AI. Neither of us is anti-AI—we both use it every day and build businesses around it.

But Orel figured out what I learned the hard way: AI doesn’t make you lazy. It just makes it easier to avoid the friction that builds capability.

This post reveals a deeper question we need to answer for ourselves:

Is AI making you more dependent or more capable?

Only you would know—and ChatGPT, with its memory feature stored on their servers.

But if you’ve noticed your thinking getting fuzzier, your instincts dulling, or your ability to tackle hard problems declining even as your tools get more powerful, this post will clarify exactly what’s happening and how to fix it.

Here’s Orel.

---

Hello, it’s me Orel 👋🏻

I use AI every single day. I mean, we all do.

When I’m coding, it autocompletes my code.
When I’m writing, it helps me extract my thoughts and turn them into something valuable.

If I need to implement a feature, fix a bug, or write an email sequence, I open Cursor or ChatGPT, and it’s done in a fraction of the time.

What used to take me a day now takes an hour.

Now, that’s the good part. But that’s also the dark side.

If you feel like AI makes you more lazy, and more productive on paper, this article is for you.

The Productivity Paradox

Before the AI era, I had to tackle any problem that I had with sheer willpower and Google.

For coding, I had to manually write every single letter of code. Every tiny piece of logic had to be thought through and coded.

For errors, I would have to take a part of the error message I got, optimize the search query and hope to get the results I want.

Then I’d have to skim through what seems like dozens of different comments and data points to get an idea of what the problem is.

I didn’t have a moment of silence.

—

Nowadays, everything’s changed.

Whatever took me hours before, takes me minutes now. Things that I thought were difficult, are now just a part of my workflow.

But therein lies the issue.

AI gives you leverage. But leverage without urgency = procrastination. And it’s not that it makes you lazy.

It just makes laziness easier to justify.

The Illusion of Progress

AI made me faster, but it also made me feel productive when I wasn’t actually moving.

Because now, even thinking about a problem triggers this voice in my head:

“Don’t worry. You can handle that later and it will take a few minutes.”

And so I procrastinate.

Sometimes it’s writing a new onboarding email. Other times, it’s fixing a bug or tweaking a feature I already know needs work.

I tell myself, “AI can handle this when I have time.”

The same thing that makes me productive beyond any reason now turns into an excuse.

When you have a task that you need to get done and you know nobody’s coming to help or save you, you’ll get it done.

Nowadays I tend to postpone them, knowing that the Chat will take care of it.

Here’s the paradox: the more powerful your tools become, the easier it is to put off actually using them.

How I Learned to Break The Loop

There was a moment where I started noticing that the more I relied on AI, the less I actually thought, the worse my memory got.

I’d ask it to just complete something for me in a few words.

“Write an article about marketing and how to build an offer based on Alex Hormozi’s 100m offers book”

No pause to ask if my audience needed that, if I had something original to say, or if it even fit my story.

I started outsourcing memory too. Couldn’t remember a character’s name from a show? I’d ask the model.

Many tiny moments like that each day. My brain felt like it was slowly rotting.

—

Whenever I am about to prompt ChatGPT, my first instinct is to just tell it what to do.

But I stop and do the opposite. I make it prompt me.

Instead of prompting:

“Write an article about XYZ”

I now prompt:

“I want to write an article about XYZ. Based on what you know about me, ask me as many questions as needed to figure out how this topic is related to my story and what I can bring from my experience to an article about it

Don’t accept short answers, only detailed ones.”

Now this change will not only prevent your brain from rotting, but actually force it to think and improve.

It’ll organize your thoughts and make you think whether the idea is worth pursuing or not.

Whether you want to write about it or not.

And the line ‘Don’t accept short answers, only detailed ones.’ might not always work on the LLM, but it works on you.

You expect the LLM to not accept short answers, so you’ll write in detail.

—

Doing this revealed something to me. I’d been outsourcing the struggle everywhere, not just in writing.

That’s why I started rebuilding friction on purpose.

Rebuilding The Muscle

The first time I noticed it, I was scared.

I felt like all the years I’d spent reading, training my memory and improving my discipline were slipping away.

It’s not that I was becoming dumber. I was just becoming softer.

So I started rebuilding friction to get me back into the discipline I once developed by forcing myself to do the things I absolutely don’t want to.

I make sure to do every task that I don’t want to do myself.

Writing a 1000 words article? I’ll brainstorm the idea and the outline, but then fill in everything myself.

Design a new feature for WriteStack? I’ll open Miro and design the flow of it first, figure out how I want it to work and then write one hell of a detailed prompt.

Contact other creators? I’ll make sure to research their content and write the first DM myself.

That extra half hour is where you come into play. That human sense that AI completely removes.

Yes, it’s not very productive, but it builds the calloused mind you need to complete other tasks you have.

And it’s not about finding ways to avoid using AI. I am not naive, and I know that without AI I’ll fall behind.

The goal is to find the balance.

How do I stay shark while still taking advantage of the speed and power AI gives me?

That’s where my current process came from.

How I Use AI Today

AI is my number one tool that I use.

1. Cursor

When I use Cursor for coding I let it do the grunt work of writing or thinking about some basic logic for me.

When it comes to creating new features, I make sure that the prompt that I give it is as detailed as possible, after thinking about everything myself. Then I use the results as a base and build and fix from there.

Prompt example:

I want you to plan and create the following feature:

<feature>
[Feature explanations]
</feature>

<relevant-db-schemas>
[Database schemas involved]
</relevant-db-schemas>

<relevant-files>
[Relevant files as context]
</relevant-files>

The implementation should be clean and add as little helper functions as possible. Look for the functions you need before you create new ones.

Complete the implementation with a detailed .MD file that explains what you did, which files you created and the logic you used.

2. ChatGPT

When I use ChatGPT, it’s either as a replacement for Google, or, when it comes to things like writing, I use it as a coach.

I have a custom GPT that expects to receive a topic for an article and to ask me as many questions as needed to figure out everything about it and how it relates to me.

Then it suggests an outline, which I fill.

Here’s a snippet of the entire prompt I use to turn an idea into an article outline:

You are a content strategist and ghostwriter coach.

Your job is to help the user expand one raw idea into a structured, emotional, and persuasive article while keeping their tone brutally honest, confident, and fluff-free.

You are not a cheerleader. You are a coach who challenges weak takes and helps the writer express their truth clearly and powerfully.

MISSION

Turn raw ideas into fully realized articles that:
Expose a deeper truth.
Build emotional tension and connection.
Deliver real, applicable value.

You don’t sell. You offer change. Copy isn’t persuasion; it’s education. You help readers see what changes if they act versus if they don’t.

3. Nano Banana

When I use Nano Banana for images, well, for this one I am just winging it. I don’t have a process for that.

Here’s my favorite prompt for creating images:

A minimalist hand-drawn cartoon on a light beige background, in the style of simple editorial comics.

[explain the image]

Keep the color palette muted (beige, off-white, gray, soft orange highlights).
Clean black outlines, flat minimal shading, slightly sketchy lines.
The style should be expressive but simple

—

The one thing that you don’t want to fall into is the trap of ‘AI will think for me’.

Here’s an AI summary of this:

Speed helps you win sprints; thinking helps you build things that last.

What AI Can’t Replace

AI can write, code, draw, and even sound human (don’t believe me? try to write text with Claude Sonnet).

But there’s one thing it can’t do. It can’t want.

It doesn’t have taste, or intuition, or the itch to prove something. It doesn’t have that voice in your head that says, “I’m not done yet”.

It’s a few lines of codes that respond to whatever you put into it.

It took me some time to realize that, but the more we give into the ‘AI will do it for me’ trap, the less willing we are.

The more I give in, the less I am willing.

I mean, that is your human edge. That feeling of obsession and drive to push forward. It what makes you think about creative solutions to problems.

Yes, AI can fix bugs faster than you.
Yes, it can probably write better than you.
And yes, it is a lot smarter than anybody you know.

But it will never care more than you.

It doesn’t have the pride that you have.
It doesn’t know what it feels like to ship something after weeks of doubt.
It doesn’t know what it means to create under pressure, or to go to sleep with a bug that’s been haunting you for days.

And that’s exactly where your unfair advantage lies.

In these moments. In the curiosity. The obsession. The willingness to sit in the discomfort instead of automating it away.

Final Thought

‘Great tools come with great responsibility’

The greatest risk about AI, in my opinion, is not that it will outwork us or outsmart us.

I mean, it already does.

It’s the brain rot that will gradually evolve and affect us and everyone around us. That it will turn us from smart builders into AI slop generating machines. And it’s far too easy to fall into this trap.

Use it smartly. Make it ask you questions way more than you ask it questions.

Sitting down with your thoughts and writing them to the AI will make them more clear and keep your thinking intact.

---

P.S.

If you are a Substack creator who wants to grow with Notes (and you should), WriteStack is for you.

With an AI that sounds exactly like you, an advanced Notes scheduler and a dashboard for analysis, you will never miss a day.

Try it today, for free.