I Broke ChatGPT's 'Robot Voice' With This 5-Second Prompt

        **Core Topic:** How to make AI (specifically ChatGPT) write in a casual, conversational, human-sounding voice instead of its default stiff, formal tone—using a single simple prompt called "Keep It Real Mode."

**Main Speaking Points:**

1. **The Problem with Default AI Writing** — Out of the box, ChatGPT sounds robotic, overly formal, and disconnected. It reads like "a dumb man's version of a rich man." It doesn't work for copy, persuasive communication, or content that needs to connect with real people.

2. **The Solution: Keep It Real Mode** — A short, simple prompt that transforms ChatGPT's output into genuinely conversational, engaging writing without needing massive complex prompts or endless back-and-forth editing.

3. **Side-by-Side Comparisons** — Multiple examples showing the dramatic difference between default ChatGPT output and "Keep It Real Mode" output on topics like schemas, subtext, a spontaneous rant about leaf blowers, and the 1953 Zurich chess tournament.

4. **Good Enough Beats Perfect** — The prompt isn't flawless (it still uses some clichés like "here's the real kicker" and too many adverbs), but it's dramatically better than default output without requiring a 60,000-word "house of cards" prompt that breaks constantly.

5. **Teaser for Future Content** — Combining voice prompts with hook prompts (based on analysis of the greatest copywriters) to take AI writing to the next level.

---

Man, if you try to write with AI, it sounds like an idiot speaking to you in the most formal presentation possible. It's a dumb man's version of a rich man. It doesn't connect. It feels impersonal. It totally sucks for copy, persuasive communication, content creation—unless you beat it with a stick like it owes you money.

So how do you make it right? Like I just talked right there—casual, informal, the kind of stuff people actually say to each other—without getting insanely complicated with 60,000-word prompts and all that madness?

I'm Jason Flatley. I've written copy that's produced hundreds of millions of dollars in sales. I'm your favorite marketer's favorite marketer because I know the real stuff. And I couldn't make ChatGPT work for me with *any* sales communication I wanted to put out there—until I discovered what I call **Keep It Real Mode**.

This is the shortest prompt I've ever used to get truly conversational copy that makes an impact without having to beat it up over and over again.

Let me show you exactly how it works.

---

### The Default ChatGPT Problem

I'm going to make this quick because I've got a date with my hot-ass wife—that's why I'm all dressed up. I don't care about fancy lights and production. Go to some Instagram marketer bro if you need that.

Here's the deal: I'll put the prompt in the description. But first, let me show you what ChatGPT looks like out of the box versus what this one simple prompt does.

Let's take a topic: **schemas**.

Most of you are probably at about a third-grade education level, so you don't know what a schema is. This might be educational. Here's what ChatGPT says by default:

> *"Imagine trying to navigate a busy city without a map. Overwhelming, right? That's what life would feel like without schemas. Schemas are mental frameworks our brain uses to organize and interpret information. They're like invisible templates built from past experience..."*

I've gotta stop right there because I know you're already about to go watch somebody get kicked in the nuts instead of this. You're bored to death. This is the kind of copy ChatGPT is renowned for.

---

### Keep It Real Mode in Action

Now let me show you what Keep It Real Mode does. I literally just said "write me an article on schemas" and then copied and pasted the prompt. That's all I gave it.

Here's what it came up with:

> *"You ever walk into a room and instantly know who's in charge? Like your brain just* gets *it before anyone even speaks? That's not intuition, bro. That's a schema.*
>
> *These little effers are running your life. Schemas are mental blueprints—tiny little programs your brain installs without asking you. They tell you how stuff should work: relationships, money, success, failure, love, pain. All of it.*
>
> *Here's the wild part: they're mostly written by your 8-year-old self. Which, let's be honest, is like giving the keys to your future to a kid hopped up on Capri Sun and emotional confusion."*

Kind of a little bit—a whole lot—better, right?

No shenanigans. Same ChatGPT window. I said "write me a short article on the power of schemas," got the drivel. Then I said the same thing but added the Keep It Real Mode prompt. Boom. Capri Sun. Emotional confusion. All of it.

---

### Another Example: Subtext

Maybe we just got lucky. Let's test it on another topic: **subtext**.

Default ChatGPT:

> *"What we say is rarely the whole story. In fact, the most powerful messages in any conversation are often the ones left unsaid. This is the realm of subtext—the hidden layer beneath our words where true intentions, emotions, and meanings live. Subtext is what turns a simple line like 'fine' into a statement loaded with disappointment, frustration, and/or resignation."*

Who the hell uses the word "resignation" when they write? Come on, ChatGPT. I'm disappointed in you.

Keep It Real Mode:

> *"You ever have someone tell you 'I'm fine' and know damn well they're not? Yeah. That's subtext. It's the quiet part of communication—the part that never makes it into the words but still punches you right in the chest.*
>
> *I used to suck at this. Like, legitimately oblivious. People would say one thing and I'd take it at face value. Didn't read the pause. The tight jaw. The 'no worries' that was clearly laced with worry.*
>
> *Turns out, what people say and what they mean? Two different effing planets."*

I didn't tell it to cuss. Maybe it's channeling some Brené Brown, Gary Vee, and Dave Chappelle. That's part of the prompt's magic.

---

### The Leaf Blower Test

Now, of course, when I'm recording this YouTube video, the leaf blower kicks up. This thing stalks me wherever I go. So I got in the moment and decided—let's just use Keep It Real Mode on the leaf blower. Spontaneous. Not planned.

Here's what it came up with:

> *"I swear to God the leaf blower has it out for me. Like, it's not even a machine anymore. It's a civilian demon with a gas engine and a grudge.*
>
> *Every time I sit down to do something important—Zoom call, deep writing session, trying to meditate and get my life together—VROOM.*
>
> *And it's never one leaf blower. No, no. It's a platoon. Like every landscaping crew in a five-mile radius got a group text: 'He's about to focus. Move in.'*
>
> *At this point I'm convinced they're following me. I've changed rooms. Closed windows. Considered faking my own death. Still..."*

All from that one Keep It Real Mode prompt.

---

### Going Left Field: The 1953 Zurich Chess Tournament

Let's go totally off the marketing/communication track. When I was 16, I decided I wanted nothing but chess books for my birthday. Trust me, I didn't get laid at all that year. My favorite book was *The Zurich International Chess Tournament 1953* by David Bronstein.

So I said: "Yo ChatGPT, write me a short article on the Zurich International Chess Tournament 1953."

Default output:

> *"The Zurich International Chess Tournament of 1953 remains one of the most legendary events in chess history. Held in Zurich, Switzerland, this candidates tournament was organized by FIDE to determine who would challenge World Champion Mikhail Botvinnik for the title. It gathered..."*

So freaking boring.

Keep It Real Mode:

> *"All right, let me be real with you for a sec. Zurich 1953 wasn't just a tournament. It was a damn war zone. Not with tanks and bullets, but with minds.*
>
> *The best chess players on the planet crammed into this pressure cooker in Switzerland, trying to claw their way to the top for a shot at the freaking world title. And it was brutal. Fifteen absolute killers. No fluff. No filler. Just raw talent, egos, and endless coffee."*

When it talks about Smyslov, who ends up winning and later beating Botvinnik:

> *"Dude was ice cold. Quiet. Classy. Moved like a ghost. Crushed like a freight train."*

All from that one simple little Keep It Real Mode prompt.

---

### It's Not Perfect—But Good Enough Beats Perfect

Let's be straight with each other. This isn't flawless. I don't love the "here's the real kicker" nonsense that LLMs like to throw in. I don't like adverbs—"legitimately," "typically"—that's the sign of a weak communicator. We *could* build a super-duper prompt filtering all that out, but then you've got a house-of-cards situation that'll probably break eventually and need constant updating.

Sometimes good enough is better than perfect.

That's the power of Keep It Real Mode. It's just this one short prompt.

---

### What's Next

The prompt is in the description. Copy it. Paste it. Use it. Whatever output you get, drop it in the comments—I'd love to see what you came up with.

And when you combine a voice prompt like this with **hook prompts**—which I've developed by analyzing the greatest copywriters who ever lived—now you're really taking it to the next level.

That'll be the next video. Or an upcoming video. I don't know, I'm insane, I don't really follow a schedule. But subscribe because I *will* create that video with those prompts for your greedy little fingers to copy and paste into ChatGPT and get some *real* stuff out of it.

Not just the nonsense everybody else is calling good.

Catch you in the next video.