This Magnifying Glass Trick Makes the AI’s Output 10X Better

We all have bad habits.

Mine? I’m an Oreo criminal.

I take the “stuff” out of one cookie and add it to another. Singles become doubles. Doubles become quads. You get the idea.

Now here’s the evil part: I put the stuff-less cookie back in the container for the next person.

(Try it. Sometimes being a little bad is enjoyable.)

So yeah, we all have bad habits.

But I’ve noticed one recently that’s become super common since ChatGPT got popular.

You could even call it a bad habit that people don’t even know about!

What is it?

It’s not editing the AI’s output after it writes. That’s a terribly BAD habit.

I’m not talking about hallucinations or fake facts. You have to check for that stuff.

What I’m referring to is editing to ensure the AI’s output is the highest quality possible.

See, here’s what nobody tells you:

AI literally cannot give you its best work on the first try.

Not because it’s lazy. Not because it’s holding back.

It’s because writing and critiquing use completely different “circuits” in its brain.

When AI writes normally (e.g., emails, blog posts, social media) it’s simply predicting what words should come next, like typing on autopilot.

But when AI critiques?

It’s comparing the text it wrote against patterns of good and bad writing. It spots weaknesses it couldn’t see before.

Both modes cause the AI to sample from different portions of its training, using different parts of its digital brain.

Now here’s the secret:

These two modes can’t run at the same time.

It’s like trying to talk and listen to yourself simultaneously, which is impossible.

So what’s the solution?

It’s something that I call the magnifying glass trick.

You force AI to write something first. Then you make it switch modes and examine its own work, like a detective.

And then?

It fixes everything that was wrong.

The result?

10X better than the original. Every single time.

Let me show you how it works.

How it Works: The Magnifying Glass Trick

This dead-simple trick makes average AI content exceptional.

You start by having the AI generate content as normal. Use your regular prompts and directions.

After it finishes writing, that’s when you apply the magnifying glass trick.

Here’s what you do:

Tell the AI to examine what it just wrote with a magnifying glass and perform an exhaustive and detailed critique of it.

Don’t ask it to fix the content. Just ask for the critique.

Why?

You’ll get a more thorough critique if the AI doesn’t have to pull double duty. It won’t try to fix the output in the same request.

Finally, once the critique is finished, you can tell the AI to fix the content. That is, if you’re satisfied with what it found.

That’s it.

Three prompts. 10X better output.

Let’s look at a specific example, so you can see for yourself.

Example of Using the Magnifying Glass Trick

Let’s try using the magnifying glass trick to help write a social media post.

Like I said before, the first step is simple. Just have the AI write your content as normal.

In this case, that’s a LinkedIn post.

AI Prompt

Write a LinkedIn post for our B2B software company’s page. We make SlideWhisper, the AI presentation software that builds slides by listening to you talk, and need to post 3x a week to stay visible.

Here’s the context: Our company Presentation Dynamics has 12,000 followers but our posts get maybe 30 likes and 3 comments – usually from our own employees. We sell to business professionals and our main buyers are Sales Directors, Marketing VPs, and Consultants. They care about looking polished, saving time, and not staying up until 2am before big presentations.

About SlideWhisper: It’s revolutionary because you literally just talk out loud about your ideas and it creates professional slides in real-time. No typing, no design skills needed. It generates custom graphics, pulls in real data from connected tools, and even remembers your presentation style. Sales teams love it because they can create customized pitches in minutes instead of hours. The AI understands context, so if you say ‘our growth story’ it knows to create a chart, not bullet points.

This week’s post topic: The nightmare of last-minute presentation requests. I want to talk about that panic when someone says ‘can you present this tomorrow?’ without making it a boring corporate lecture. Our software helps with this but I don’t want to be too salesy.

Goals for the post:
– Get actual engagement (comments, not just mercy likes from coworkers)
Sound human, not like a corporate robot.
– Position us as understanding real presentation panic.
– Keep it short enough people will actually read it (under 150 words).
– It should feel like it’s written by someone who’s actually been ambushed with surprise presentation requests, not by marketing people who’ve never had to create 20 slides by tomorrow morning.
– Don’t use emoji or hashtag spam because our audience hates that stuff.
– Don’t use emdashes which are a telltale sign of AI, but you can use some bullet points if that is useful.

AI Response

“Can you present this tomorrow?”

Five words that turn your Tuesday into a caffeine marathon.

You know the drill. It’s 4:47pm. You’ve got maybe 20 slides to create. Your “quick dinner” becomes whatever’s in the vending machine. By 11pm, you’re debating whether slide 12 needs more blue. By 2am, you’re googling “professional synonyms for synergy” while your third energy drink judges you silently.

The worst part? You KNOW your content. You could explain this stuff brilliantly over coffee. But somehow translating that into slides feels like solving a puzzle where half the pieces are missing and the picture keeps changing.

Here’s what changed everything for our team: we stopped typing and started talking. SlideWhisper listens while you explain your ideas out loud, then builds the slides. Like having a designer who actually gets what you mean when you say “make it pop, but professional.”

No more 2am font debates. Just you, talking through your brilliance, getting slides that actually look like you know what you’re doing.

Because you do.

I think the result followed our prompt pretty well.

It wrote a post that gets what presentation panic really feels like. There aren’t any emojis or hashtags. It’s also missing those em dashes that scream AI.

But something’s off.

It has that salesy “feel” to it. I can’t quite put my finger on it.

What do you think?

So let’s have the AI use the magnifying glass trick and see what happens.

AI Prompt

Now, examine what you just wrote with a magnifying glass and perform an exhaustive and detailed critique of it.

AI Response

Exhaustive Critique of the SlideWhisper LinkedIn Post

What Actually Works

The Opening Hook: “Can you present this tomorrow?” is genuinely strong. It’s a visceral trigger that immediately puts readers in that exact panic moment. No throat-clearing, no setup needed.

Specific Time Details: “4:47pm” is better than “late afternoon” because it feels real. The “11pm debating blue” and “2am googling synonyms” progression effectively builds the nightmare timeline.

The Energy Drink Detail: This lands well. It’s specific, relatable, and has personality without trying too hard.

Critical Failures

Word Count Disaster:
The post is approximately 180-190 words, blowing past the 150-word limit by 20-30%. This breaks a core requirement and shows poor editing discipline.

The “You KNOW your content” Paragraph: This is the weakest section. The puzzle metaphor is clichéd and adds nothing. “Translating into slides” is corporate-speak masquerading as casual language.

Sudden Sales Pivot: The transition to SlideWhisper feels jarring. We go from commiserating about pain to suddenly “Here’s what changed everything for our team” which screams marketing script. Nobody talks like this.

“Make it pop, but professional”: This attempt at insider humor falls flat. It’s trying too hard to be relatable and ends up feeling forced.

The Ending: “Because you do” is attempting to be punchy but comes off as cheesy motivational poster territory. It’s the kind of false empowerment that makes people scroll away.

Tone Problems

Inconsistent Voice: The post can’t decide if it’s bitter comedy (energy drinks judging you), practical advice (explaining SlideWhisper), or motivational speaking (you’re brilliant!). Pick one lane.

“Our team” Problem: This immediately signals “marketing department wrote this” rather than “actual human who uses SlideWhisper.” It breaks the peer-to-peer authenticity.

Missing Engagement Drivers

No Question:
The post doesn’t ask anything. There’s no natural comment prompt. People need an excuse to engage that doesn’t feel like helping a company’s marketing metrics.

No Controversy: Everything is too safe. There’s nothing here that would make someone go “FINALLY someone said it” or “Actually, I disagree because…”

No Vulnerability: Real vulnerability would be admitting a specific failure, not just generic “we’ve all been there” platitudes.

The Product Integration

Too Much Feature Description: “Listens while you explain your ideas out loud, then builds the slides” is a feature description, not a benefit story. We’re telling, not showing.

Weak Metaphor: “Like having a designer who actually gets what you mean” doesn’t capture the revolutionary aspect strongly enough.

What Should Have Been Cut
– The entire puzzle metaphor paragraph
– “Because you do” ending
– At least one of the late-night timeline details
– The “make it pop” line

What’s Missing
– A specific, embarrassing presentation failure story
– Any mention of the actual outcome (what happened after using SlideWhisper?)
– A natural conversation starter
– Industry-specific pain (Sales Directors losing deals, consultants looking unprepared)

The Verdict
This post is trying to check boxes rather than actually connect. It has moments of authentic voice (energy drink, 4:47pm) but drowns them in marketing-speak and forced relatability. It would likely get the same 30 mercy likes because it doesn’t give anyone a real reason to engage. It’s neither vulnerable enough to create connection nor provocative enough to start discussion. It’s the presentation equivalent of slide 12 that you’re not sure needs more blue—technically fine but ultimately forgettable.

The most damning critique: If this showed up in my LinkedIn feed, I’d scroll past it in about 2 seconds because nothing in the first two lines makes me need to read more. It confirms what I already know without adding new insight or entertainment value.

Woah.

That’s quite a critique!

Now at this point, you can just tell the AI to rewrite it based on the critique.

But what if there’s something in the critique you don’t like?

Simple.

Just tell the AI not to implement those specific things.

Let’s see the result.

AI Prompt

Please rewrite what you just wrote based on your critique.

AI Response

“Can you present this tomorrow?”

My VP asked at 4:47pm yesterday. Board wants to see Q1 pipeline analysis. Twenty slides.
Here’s what I didn’t do: Start building slides.

Here’s what I did: Spent 40 minutes choosing between “modern” and “professional” templates. Another hour finding the exact shade of blue from our brand guide. Two hours making bullets align perfectly. By midnight, I had 4 slides and a stress headache.

The stupid part? I know our Q1 pipeline cold. Every deal, every blocker, every opportunity. Could’ve pitched it perfectly over coffee.

3am revelation: I’m not bad at presenting. I’m bad at PowerPoint.

So here’s what’s changed: Now I just talk through my presentations to SlideWhisper. I explain the pipeline story out loud, it builds the slides while I’m talking. Yesterday’s emergency deck? 35 minutes instead of my usual all-nighter. Actually looked professional.

What’s the dumbest thing you’ve spent hours on instead of actual presentation content? Mine was whether to italicize a word. On slide 23. That nobody saw.

I’m no LinkedIn expert, but that sounds great!

Now here’s another trick.

If you’re not happy with the AI’s result, just tell it. Then have it rewrite everything based on your feedback.

Here’s why this magnifying glass trick is so powerful.

It helps reveal things you might not have thought of. You’ll notice new ideas only after reading the AI’s original output.

Advantages of the Magnifying Glass Trick

Here’s some of the advantages of implementing the magnifying glass trick:

  • Catches blind spots you’d never see yourself The AI finds problems that you might not have noticed. It’s like having fresh eyes every time.
  • Transforms generic garbage into standout content First drafts sound like everyone else. The magnifying glass finds what makes yours forgettable and then fixes it.
  • Saves you from sending embarrassing mistakes How many times have you sent something, then noticed the obvious flaw? The magnifying glass catches it first.
  • Works with any type of content Emails, ads, landing pages, proposals, social posts. If it has words, the magnifying glass improves it.
  • Takes just 60 seconds extra Three prompts instead of one. That’s it. For 10X better results.
  • Builds your editing instincts Watch the AI critique enough times? You start spotting these problems yourself.
  • Eliminates writer’s guilt You know that nagging feeling that your copy could be better? The magnifying glass kills it.

When to Use the Magnifying Glass Trick

Not every piece of content needs this treatment. But these situations? The magnifying glass is non-negotiable:

  • Before sending an email to your list Your subscribers’ trust is gold. One boring email can lose them forever.
  • When writing sales copy The difference between 2% and 4% conversion? That’s double your revenue. Worth the extra minute.
  • For important client communication Proposals, project updates, difficult conversations. When it matters, use the magnifying glass.
  • Creating social media ads You’re paying for every impression. Bad copy is literally burning money.
  • Landing page headlines and copy First impressions determine bounce rates. The magnifying glass ensures yours hook instantly.
  • Cold outreach messages You get one shot. Make it count.
  • When something feels “off” but you can’t pinpoint why That gut feeling? The magnifying glass finds what’s causing it.
  • Before publishing blog posts or articles Your content represents your expertise. Generic posts damage your authority.
  • Any time you’re copying/pasting templates Templates are starting points. The magnifying glass makes them yours.

Here’s the rule:

  • If someone’s going to judge you or your business by this content? Use the magnifying glass.
  • If it affects revenue, reputation, or relationships? Use the magnifying glass.
  • If you spent more than 5 minutes writing it? Spend 1 more minute with the magnifying glass.
Previous Lesson
Next Lesson