When was the last time you disagreed with your AI?

Hi, Gopi here.

I’ll be honest.
Writing has never been my thing.

Ever since school, my English teachers would berate me for - well, just sucking at English.

I never got it. I could speak fine in class, but the moment a pen hit the paper, it was like my brain and hand were in a long-distance relationship.

Looking back, the problem was obvious.
Thinking.

My brain would be racing at 100km/h, but my thoughts just wouldn't make it onto the page.

Everyone said, "Just proofread."
Simple, right?

But in an exam, you’re not just writing. You’re calculating time, wondering if you wrote your name down, and praying your mum packed the right sandwich (a vital thought for lil Gopi).

Somehow, I persevered. Learned to write (mediocre at best), and passed.

But now I have a new challenge: writing with AI.

Ever since ChatGPT went mainstream, I’ve been ruthless with it. As an AI Engineer, it's kinda my job.
I honestly don't know the last time I wrote an email without it.

And that’s alarming.

Why?
Because I stopped thinking.

The one thing that held me back in school is the very thing I'm now desperate to get back.

Using AI saves time, sure. But it was costing me the one thing a founder needs most: the ability to think critically.

I learned this the hard way.

Asking AI to "brainstorm" with me was innocent at first. Then I got lazy and just asked it to do the work for me.

This got really bad when I was doing outreach for my startup, Hoom.
It felt so efficient. Generate an email and boom! ship it.

The result? Silence.

The emails were perfect. That was the problem.
They sounded like everyone else. A complete loss of originality.

And it didn't stop there.

Asking AI for help with product design, code, or copywriting gave me the same generic, plausible-sounding advice.
It was a feedback loop of confusion. The AI just gives you the answer it thinks you want to hear.

Outsourcing your thinking to AI is the fastest way to kill an idea.

And that’s where I had to break the loop.

I decided to make AI my biggest enemy. My opposition. My adversary.

It sounds counterintuitive, but I started forcing AI to criticize my ideas. To find flaws, inconsistencies, and gaps in my thinking.

And that's when AI, the source of my problem, became my solution.

By presenting me with different, critical perspectives, it forced me to think deeply about the problem again. It reduced my dependency and actually rebuilt my critical thinking skills.

Suddenly, things started working.
Better responses to emails. More coherent plans. Finding vulnerabilities before they could create chaos.

Yes, it's more painful to write and think for yourself again.

But it helped me find my voice.
It gave me my brain back.

Actionable Takeaways

• Next time you use AI, ask it to poke holes in your plan.

• Rewrite the AI's output in your own words. Don't copy-paste.

• Compare your thinking vs. AI’s. Where do you disagree? Why?

• Use AI for opposition, not just for answers.

Links

As an AI Engineer, people often ask me how I approach critical analysis. I have developed a platform to analyze my own texts:

• AI Writing Critiquer

But interested on the prompting? Here are a few resources to get you started on your next world-changing idea:

• 5 ChatGPT Prompts To Use Criticism To Grow Your Business

• Prompt Engineering - Constructive Critic Prompts

• The Constructive Critic Prompt Framework