I Trained AI to Write Like Me (And It Changed How I Work)

Everyone's freaking out about AI replacing copywriters.

I did something different.

I transferred myself into AI.

Not metaphorically. Literally. I fed it everything I'd ever written – emails, ad copy, landing pages, campaign briefs – and trained it to sound exactly like me.

Why? Because I saw what was coming before most people did.

While marketers were debating whether AI was a threat, I was turning it into my unfair advantage.

Here's what actually happened.

The Early Days (When Everyone Was Still Skeptical)

This was early. Before everyone had ChatGPT accounts. Before "AI copywriter" became a job title.

I was working across four countries. Writing copy for Dubai audiences in the morning. Singapore campaigns in the afternoon. Melbourne clients by evening.

Different markets. Different tones. Different cultural nuances.

And I was burning out.

That's when I realized: I'm doing the same process over and over. Research. Brief. First draft. Revisions. The mechanics were repetitive. The thinking wasn't.

So I asked myself: What if AI could handle the mechanics while I focused on the strategy?

Not replace me. Amplify me.

How I Actually Did It

Here's the process I used to train AI to write in my voice:

Step 1: I collected everything I'd written

• 200+ email campaigns

• 50+ landing pages

• Hundreds of ad variations

• Campaign briefs

• Social copy

• Client presentations

Everything.

Step 2: I organized it by type and performance

I didn't just dump it all into AI. I separated:

• High-converting copy (what actually worked)

• Medium performers (decent but not great)

• Failures (what bombed and why)

The failures were just as important. AI needed to know what NOT to do.

Step 3: I created my voice document

This wasn't just "write like me." I got specific:

• Sentence structure I use (short. punchy. varied length)

• Words I love (actually, here's the thing, look)

• Words I avoid (leverage, utilize, synergy)

• How I open emails (direct, no fluff)

• How I handle objections (address them head-on)

• My rhythm (mix of long and short sentences)

I wrote out the rules of "my voice."

Step 4: I trained it with examples

I'd give AI a prompt:

"Write a welcome email for a B2B SaaS product targeting marketing managers in Singapore"

Then I'd compare its output to how I'd actually write it.

Wrong tone? I'd show it why.

Wrong structure? I'd rewrite it and explain the difference.

Right direction? I'd refine it further.

Over and over. Hundreds of iterations.

Step 5: I created custom prompts for different scenarios

I built a library of prompts for:

• Welcome sequences

• Launch campaigns

• Objection-handling emails

• Ad copy (by platform)

• Landing page headlines

• Subject lines

Each prompt included:

• My voice guidelines

• The specific context

• Examples of past successful copy

• What to avoid

What Changed (The Results Nobody Talks About)

Here's what actually happened when I started working this way:

Speed increased 3x

What used to take 2 hours (research, write, revise) now takes 30 minutes.

AI handles the first draft. I handle the strategy and refinement.

Quality stayed the same (sometimes better)

This surprised me.

AI trained on my best work started catching patterns I didn't consciously know I used. It made connections I hadn't thought of.

Sometimes its first draft was better than mine would've been.

I could scale across markets

Remember: I was working across four countries with completely different audiences.

I trained AI variations for each market:

• Dubai voice (more formal, relationship-focused)

• Singapore voice (data-driven, credibility-heavy)

• Kazakhstan voice (transformation-focused, accessible)

• Sydney & Melbourne voice (casual, direct, authentic)

Now I could switch between them in seconds.

My role evolved

I stopped being a copywriter who thinks about strategy.

I became a strategist who happens to write copy.

Big difference.

The Things AI Still Can't Do (And Probably Never Will)

Let me be clear: AI didn't replace me. It's a tool, not a replacement.

Here's what it still can't do:

1. Strategic thinking

AI can write the campaign. It can't decide if you should run the campaign.

Should we launch now or wait? Email or ads? What's the actual goal here?

That's human work.

2. Understanding nuance

Working across different markets taught me this: cultural context matters.

What's persuasive in Dubai doesn't work in Singapore. What converts in Kazakhstan needs adjustment for Melbourne.

AI knows patterns. It doesn't understand context the way humans do.

3. Reading the room

Is the client ready for a bold approach? Will this message land well given current events?

AI can't read subtext. It can't sense tension. It can't feel when something's off.

4. Original strategic insights

AI remixes what exists. It doesn't create genuinely new frameworks or approaches.

Every original strategy, framework, or positioning I've created? That came from human experience, not AI.

5. Client relationships

People hire people, not machines.

AI can write the pitch. It can't build trust over coffee. It can't read body language. It can't navigate politics.

That's still human territory.

How This Actually Works in Practice

Here's my workflow now:

Step 1: I do the strategic thinking (human)

• What's the goal?

• Who's the audience?

• What's the big idea?

• What channels make sense?

• What's the core message?

Step 2: AI creates first drafts (AI)

• I feed it my strategy

• It generates variations based on my voice

• Multiple options, different angles

Step 3: I refine and adapt (human)

• Pick the best direction

• Adjust for context AI missed

• Add personality and nuance

• Make it sing

Step 4: AI handles variations (AI)

• Need 10 subject line options? Done.

• Need ad copy for 5 audience segments? Done.

• Need this email in 3 different tones? Done.

Step 5: I make final calls (human)

• Which version to test?

• What to kill?

• What to push forward?

The AI handles volume. I handle judgment.

The Mistake Everyone Makes with AI

Most people use AI wrong.

They treat it like a magic button:

• "Write me a welcome email"

• "Create ad copy for my product"

• "Generate a landing page"

Then they copy-paste whatever it spits out.

That's not using AI. That's being lazy.

Here's the difference:

Bad approach:

"Write a welcome email for my marketing blog"

My approach:

"Write a welcome email for my marketing blog targeting early-career marketers in Melbourne who want to break into strategic roles. Use my voice guidelines [attached]. Similar to the tone in [example email]. Focus on credibility from my cross-market experience. Keep it under 150 words. No generic motivational language. Direct CTA to read the latest post about email sequences."

See the difference?

One gives AI context, constraints, and direction.

The other hopes AI can read your mind.

What Being Early Actually Meant

I jumped on AI before it was cool. Before everyone had "AI-powered" in their LinkedIn headline.

Here's what that gave me:

1. Time to experiment

While others were still deciding if AI was worth learning, I had 6 months of testing and refinement.

I made mistakes when the stakes were low. I learned what worked when nobody was watching.

2. A real competitive advantage

By the time everyone else started using AI, I had custom systems built.

They were copying prompts from Twitter. I had a trained AI that wrote in my exact voice.

3. Credibility

Being early means you can say "I've been doing this since..." when everyone else just started yesterday.

That matters when you're job hunting or pitching clients.

4. Pattern recognition

I saw what worked. I saw what was hype.

I could separate the useful tools from the noise because I'd actually tested them.

The Real Skill Now (What Actually Matters)

Here's what I've learned:

The future isn't "AI vs. humans."

It's "humans who use AI vs. humans who don't."

And the skill that matters isn't writing anymore.

It's knowing what good writing looks like.

You need to be good enough to:

• Evaluate AI's output

• Know when it's off

• Understand why it's wrong

• Fix it quickly

• Direct it toward better results

AI can generate 100 variations. You need to know which one will actually convert.

That's the skill.

What This Means for Copywriters (And Marketers)

If you're scared AI will replace you, you're asking the wrong question.

The question isn't "Will AI replace copywriters?"

It's "Will copywriters who use AI replace copywriters who don't?"

Yes. Obviously yes.

Here's my advice:

1. Learn it now

Not next month. Not when you have time. Now.

The gap between early adopters and late adopters is getting wider every day.

2. Train it on your work

Don't use generic AI. Build YOUR version of AI.

Feed it your best work. Teach it your voice. Make it yours.

3. Focus on the strategy

Let AI handle the execution. You handle the thinking.

What should we say? To whom? Why? When?

That's your job now.

4. Stay human

AI can write like you. It can't think like you.

Your experiences, your insights, your strategic judgment – that's irreplaceable.

5. Move fast

This is changing quickly. What works today might not work next month.

Stay curious. Keep testing. Adapt faster than everyone else.

The Bottom Line

I didn't transfer myself into AI because I thought copywriting was dead.

I did it because I saw where things were going.

AI isn't replacing good marketers. It's making good marketers unstoppable.

The question is: Are you using it to get better, or are you pretending it doesn't exist?

I made my choice early.

What's yours?

Want to know the exact prompts I use? Hit me up. I'm building a library of the frameworks that actually work.

thezhanakhmetalim@gmail.com linkedin.com/in/alimzhanakhmet

P.S. This post? Written with AI. Then refined by me. That's the point.