The "Generic AI" Trap: How to Beat Applicants Who Use ChatGPT

ChatGPT is trained to be average. That's exactly how it sounds.

January 30, 2026 · 10 min read

I can spot a ChatGPT essay in about 15 seconds.

It's not subtle. There are tells, specific patterns that scream "this person pasted the prompt and clicked generate."

And if I can spot them, so can admissions officers who read 10,000 essays a year.

If you're going to use AI (and you should), you need to use it better than everyone else.

The AI-Speak Warning Signs

When I read a ChatGPT-generated essay, these patterns jump out immediately:

1. The Em Dash Epidemic

ChatGPT loves em dashes. "I learned that leadership—true leadership—requires more than just vision." Every AI essay has three of these per paragraph.

2. The "It's Not X, It's Y" Construction

"It's not about working harder. It's about working smarter." This formula appears in almost every AI-generated essay because it sounds insightful without saying anything specific.

3. Excessive Signposting

"First, I realized... Second, I learned... Third, I discovered..." AI loves numbered lists and obvious transitions. Real writing doesn't announce its structure that loudly.

4. Strange Verbiage

Words like "delve," "spearhead," "synergize," "leverage" (as a verb), and "transformative" appear way more often in AI text than in human writing. They're corporate filler.

5. Overly Concise, Perfect Sentences

Real human writing has hesitation, run-on sentences, and imperfect phrasing. AI text is suspiciously polished: every sentence the same rhythm, every paragraph the same length.

"The moment I see these patterns, my thought is: Is there actually signal here, or should I stop reading?"

This doesn't mean I immediately reject the content. But my skepticism goes up. And if I'm skeptical in 15 seconds, imagine what happens with an overworked admissions reader.

The Real Problem: Prompt-First Thinking

Most applicants use AI backwards.

They paste the essay prompt into ChatGPT and ask for a draft. The AI generates 400 words of generic advice dressed up in first-person. The applicant tweaks a few sentences, swaps in their company name, and submits.

This is the worst possible workflow.

ChatGPT generates from patterns. It gives you the average of everything it's seen. That's literally what it's trained to do. So when you ask it to write your leadership essay, you get the average leadership essay, the same one 10,000 other applicants are submitting.

Prompt-First vs. Story-First: A Side-by-Side Comparison

Let me show you the difference concretely.

The Prompt: "Describe a time you demonstrated leadership."

Prompt-First Approach (ChatGPT)

You paste the prompt and type: "Write a 500-word MBA essay about a time I demonstrated leadership. I'm a product manager at a fintech startup."

ChatGPT produces:

"As a product manager at a leading fintech startup, I have had numerous opportunities to demonstrate leadership. One particularly memorable experience occurred when our team faced a critical deadline for launching our flagship payments product. Recognizing the need for decisive action, I spearheaded a cross-functional initiative that brought together engineering, design, and marketing stakeholders. Through effective communication and strategic coordination, we successfully launched the product on time, resulting in a 40% increase in user engagement. This experience taught me that true leadership is not about authority—it's about inspiring others to achieve a common vision."

This essay could be written by any PM at any startup. It has no texture, no friction, no proof.

Story-First Approach (ChatMBA)

Before you ever see the prompt, ChatMBA interviews you:

"Tell me about a product launch that almost failed."

"Our payments integration was three weeks behind. Engineering didn't trust the vendor."

"Why didn't they trust the vendor?"

"The vendor had missed two milestones. Our lead engineer called them 'incompetent' in a team meeting."

"What did you do in the 24 hours after that meeting?"

"I called the vendor's PM directly. We rebuilt the timeline together. I presented it as our joint plan."

"How did your lead engineer react when you presented the new plan?"

"He was skeptical. He asked me if I was 'covering for' the vendor. I told him I was covering for us."

Now you have an essay:

"Three weeks before launch, our lead engineer stood up in standup and called our payments vendor 'incompetent.' He wasn't wrong—they'd missed two milestones. But the product wasn't going to ship without them. That night, I called the vendor's PM directly. We rebuilt the timeline together. The next morning, I presented it—not as the vendor's recovery plan, but as ours. When my engineer asked if I was 'covering for' the vendor, I told him I was covering for us. The product shipped on time."

Same underlying experience. Completely different essay. The Story-First version has specifics, dialogue, and character. It couldn't be written by anyone else.

The Story-First Approach

How to use AI without sounding like everyone else:

Step 1: Build Your Story Bank First

Don't ask AI to write. Ask it to interview you.

Before you generate a single draft, use AI to extract your stories:

  • "What's a time you almost quit something? Why did you stay?"
  • "Who was the most difficult person you worked with? What did you learn?"
  • "What's a skill you're proud of that doesn't show up on your resume?"

Save the answers. That's your raw material.

Step 2: Generate Options, Not Drafts

When you're ready to approach an essay, don't ask for "a draft." Ask for:

  • "Give me 10 different angles for answering this prompt"
  • "What are three unconventional ways to frame a leadership story?"
  • "Here are three stories from my bank. Which one fits this prompt best, and why?"

Flip through 20–30 options before committing to one.

Step 3: Write It Yourself (With Structure Help)

Once you've chosen an angle, draft it yourself. Use AI for:

  • "Help me structure this into a three-part arc"
  • "This paragraph feels weak. How can I tighten it?"
  • "Is there a more specific way to say 'I learned about teamwork'?"

Your voice stays in the essay. AI helps you sharpen it.

Step 4: Iterate on Sentences, Not Essays

Do not regenerate entire drafts. Work at the sentence level:

  • "Tweak this sentence. I'm trying to convey urgency without sounding dramatic."
  • "Give me 5 alternatives for this opening line."
  • "This ending is flat. Give me three options for a stronger close."

The essay stays yours because you assembled every piece.

What ChatMBA Does Differently

Generic AI is exactly that: generic. It doesn't know what Stanford wants vs. what Booth wants. It doesn't know that "spearhead" triggers eye-rolls. It doesn't remember your previous stories.

ChatMBA is built for this specific workflow:

  • Persistent memory: Your Story Bank carries across sessions
  • School-specific intelligence: It knows what each program values
  • Interview mode: It extracts stories through questions
  • Anti-pattern detection: It flags "AI-speak" in your drafts so you can fix them

The Bottom Line

Everyone is using AI now. The question is how effectively you use it. Merely using it is no longer a differentiator.

The applicants who stand out will use AI to find their stories faster. The essay should still sound like you, just a sharper, more excavated version of you. (For the case on why AI is like a calculator, see our companion piece.)

Stop pasting prompts. Start building a Story Bank.

Ready to find your story?

Stop staring at the blank page. Let's start excavating.