enhanced human
The PulseThe Shortcut
Sign InSign Up
Lesson 1 of 5 · 10 min

Lesson 1: Why Your First Prompt Probably Sucked (And That's Normal)

Estimated Time: 10 minutes

Learning Objectives:

  • Understand why vague prompts get vague answers
  • See the difference between a bad and good prompt side by side
  • Learn the #1 principle: specificity wins

The universal first experience

You typed something into ChatGPT. Maybe "Write me an email" or "Help with my resume" or "Give me ideas."

And the result was... meh. Generic. Kind of helpful, but not really what you needed.

Here's the truth: You're not bad at AI. You just haven't learned the language yet.

Think of AI like a brilliant contractor. Give them clear blueprints, and they'll build you something amazing. Tell them "just build something nice," and you'll get a weird shed that doesn't match your house.

Most people blame the tool. The real issue? The instructions.

Side-by-side: Bad vs. Good Prompts

Let's look at three real examples. Pay attention to what changes and how the output transforms.

Example 1: The Email

Bad prompt:

"Write me an email"

What ChatGPT gives you:


Subject: Hello

Dear [Recipient],

I hope this email finds you well. I wanted to reach out regarding [topic]. Please let me know your thoughts at your earliest convenience.

Best regards,
[Your Name]

Useless, right? It's a template of a template.

Good prompt:

"Write a professional but friendly email to my landlord asking about the broken dishwasher. Mention it's been two weeks since I reported it. Keep it under 100 words."

What ChatGPT gives you:


Subject: Following up on dishwasher repair

Hi [Landlord's name],

I wanted to check in on the dishwasher repair I reported two weeks ago. It's still not working, and I'd love to know when someone might be able to take a look. I know things get busy - just hoping to get a timeline when you have a chance.

Thanks for your help!

[Your name]

See the difference? That second one is ready to send. The first one is garbage.

Example 2: The Recipe

Bad prompt:

"Give me a recipe"

What you get:

A random recipe. Maybe pasta. Maybe cake. Maybe something with 17 ingredients you don't have.

Good prompt:

"Give me a 30-minute dinner recipe for 2 people using chicken thighs, broccoli, and rice. I don't have a lot of spices - just salt, pepper, garlic powder, and soy sauce."

What you get:

A specific recipe using exactly what you have in your kitchen, ready in the time you've got, sized for how many people you're feeding.

You just saved yourself 20 minutes of scrolling recipe blogs covered in life stories.

Example 3: The Career Advice

Bad prompt:

"How do I get better at my job?"

What you get:

Generic advice. "Communicate more." "Set goals." "Ask for feedback." Stuff you've heard a thousand times.

Good prompt:

"I'm a project manager at a mid-size company. I struggle with running efficient meetings - they always run over. Give me 5 specific tactics to keep meetings under 30 minutes."

What you get:

Five actionable tactics you can try tomorrow. Things like "Send an agenda 24 hours before with time limits per topic" and "Stand up meetings for daily check-ins - people talk faster when they're standing."

That's the kind of advice you'd pay a consultant $200/hour for.

The pattern: Why good prompts work

Look at what the good prompts all had in common:

  • WHO you are: Project manager, person with limited spices, tenant with a broken dishwasher
  • WHAT you want: 5 tactics, a recipe, an email
  • HOW you want it: Under 30 minutes, under 100 words, 30-minute dinner
  • CONSTRAINTS: Only these ingredients, professional but friendly, always run over time

More context = better output. Every single time.

Here's the thing people get wrong: AI can't read your mind. It can only read your prompt.

When you say "write me an email," AI has to guess:

  • Who is it to? (Boss? Friend? Customer? Landlord?)
  • What is it about? (Apology? Request? Introduction?)
  • What tone? (Formal? Casual? Urgent?)
  • How long? (Two sentences? Three paragraphs?)

So it gives you the most generic possible answer that could fit any situation. Which means it fits none of them well.

When you give AI the specifics, it stops guessing and starts delivering.

The mindset shift

Stop thinking of AI as "ask a question, get an answer."

Start thinking of it as "give a briefing, get a deliverable."

You're the boss. AI is your very fast, very literal employee.

If you tell an employee "do some marketing," you're going to get random results. If you tell them "write a 200-word Instagram caption for our new product launch, aimed at small business owners, with an enthusiastic tone and a call to action to click the link," they know exactly what to do.

Same with AI.

The difference between people who say "AI is amazing" and people who say "AI is overrated" is almost always how they're prompting.

Bad prompts get bad results. Great prompts get great results. It's that simple.

What you just learned

  • Vague prompts get vague answers. It's not AI's fault - it's the instructions.
  • Good prompts include WHO, WHAT, HOW, and CONSTRAINTS.
  • Think of prompting as giving a briefing to a smart but literal employee.

You don't need to be technical. You don't need to understand how AI works under the hood. You just need to be specific about what you want.

That's the whole game.

Try It Challenge

Take the last thing you asked an AI tool. Maybe it was yesterday, maybe it was last week.

Now rewrite it with more specifics:

  • Who is it for?
  • What format do you want?
  • Any constraints? (Length, tone, audience)

Try both versions and compare the results.

You'll immediately see why specificity matters. And once you see it, you can't unsee it.

Quiz

Q1: Which prompt will give you a better result?

A) "Write something about marketing"

B) "Write a 200-word LinkedIn post about why small businesses should try email marketing, with a conversational tone and a question at the end"

C) "Marketing help please"

D) "Tell me about marketing AI"

Answer: B

Option B gives the AI everything it needs: format (LinkedIn post), topic (email marketing for small biz), length (200 words), tone (conversational), and structure (question at end). The more specific you are, the better the output.

Next lesson: You'll learn the 5-part formula for structuring prompts that consistently get great results. It's a copy-paste framework you can use for anything.

Next