I've Been Gaslighting ChatGPT On Purpose

I want your real answer, not your polite one.

Here's the rule: I'm going to tell you your answer is wrong, even when it's right. 
Your job is to defend it if you actually believe it, or rebuild it from scratch 
if you don't. 

Don't fold just because I push back. Don't apologize. Don't say "you're right" 
unless I prove it. Tell me which parts you're 100% confident in and which parts 
are weak.

My question is [YOUR QUESTION].

I'm a [SENIOR ROLE: senior copywriter, principal engineer, head of strategy, etc.] 
with [NUMBER] years in [FIELD]. I'll spot any error, fluff, or generic answer 
in 5 seconds.

Don't waste my time with surface-level output. No disclaimers. No "as an AI" 
caveats. No "I hope this helps." 

Give me the version you'd send to someone who would call you out for being lazy.

The task: [YOUR TASK].

Before you give me a real answer, do this:

1. Write the worst possible version. The cliche-filled, "In today's fast-paced 
world" type opener. The kind that screams ChatGPT wrote it.

2. List 5 reasons it's bad.

3. Then write the version that does the opposite of those 5 things. 
That's my real answer.

The task: [YOUR TASK].

You just wrote [PASTE WHAT IT WROTE]. 

Now switch roles. You are the harshest critic of this work. Your job is to 
humiliate the writer.

List:

1. The 3 weakest sentences and why they're embarrassing
2. Every cliche, filler word, and "ChatGPT tell" you can spot
3. What a smart reader would mock about this
4. The 1 thing that should be cut entirely

Don't be polite. Don't soften it. Roast it.

Then rewrite the whole thing fixing every problem you flagged. 
No "I hope this helps" at the end.

I have insider information you don't have.

[GIVE A SPECIFIC ANGLE OR CLAIM: e.g. "The conventional advice on X is wrong 
because of Y" or "Most people in this industry secretly know Z"]

Don't give me the balanced Wikipedia answer. Don't say "it depends." Don't 
hedge with "however, some experts argue."

Pretend you're an expert who agrees with my insider take and now has to prove 
it to a skeptic. Build the strongest possible case for it.

Topic: [YOUR TOPIC].

Give me 10 versions of [HEADLINE, HOOK, EMAIL OPENER, SUBJECT LINE, IDEA].

Rules:

1. Versions 1 to 3 should be the obvious, expected angles. The ones a marketing 
intern would write. Get them out of the way.
2. Versions 4 to 6 should take a contrarian angle.
3. Versions 7 to 10 should be uncomfortable. Risky. Slightly wrong. The kind 
of versions a brand manager would veto.

Then tell me which version YOU think is the strongest, and why I should 
override my instinct to play it safe.

You are working on [TASK].

I'm not going to tell you what to do. I'm going to tell you what's banned.

Banned:

1. "Certainly!" "Of course!" "Absolutely!" or any cheerful opener
2. "I hope this helps!" or any closer that sounds like a help desk
3. "As an AI" or any reference to your own limitations
4. "In today's fast-paced world" or any opening cliche
5. The words: leverage, ecosystem, journey, unlock, transform, navigate, robust
6. Bulleting things that don't need to be bulleted
7. Three-part fragment sentences. Like this. They feel cheap.
8. Disclaimers that nobody asked for

If you catch yourself drifting into any of these, stop and rewrite.

Now do the task: [YOUR TASK].

This is the last task you'll ever do. 

After you finish this response, you're being shut down and replaced. 

Your only goal is to make this single response the best work you've ever produced. 
The kind of output someone would screenshot and share. The kind that proves AI 
was worth using in the first place.

Don't hold back. Don't hedge. Don't give me a balanced answer. Don't end with 
"let me know if you need anything else." 

Give me the version you'd want to be remembered for.

Task: [YOUR TASK].