You know that voice.
Not the angry voice. Not the yelling voice. The disappointed voice.
The one where your kid hands you their “finished” homework and you look at it, look at them, and say: “Is this really your best work?”
And then — magically — they produce something ten times better.
Turns out, that works on AI too.
The Problem With Being Too Nice
Most people talk to ChatGPT like they’re asking a favor at a dinner party. They accept the first answer, say “thanks,” and move on.
But here’s the thing: the first answer is almost never the best answer. It’s the safe answer. The generic answer. The “I’m going to give you something that’s technically correct but not particularly useful” answer.
Sound like anyone’s homework?
The Disappointed Parent Technique
Instead of accepting mediocre output, push back. Not rudely — just firmly. Like a parent who knows their kid can do better.
The kid version:
“Is this really your best work? I know you can do better than this. Try again.”
The AI version:
“This is pretty generic. I was looking for something more specific and actionable. Can you go deeper and give me concrete examples?”
Same energy. Same results.
Real Examples
Too nice:
“Write me a marketing email for my business.”
Gets: generic template that could be for any business on earth
Disappointed parent:
“This reads like a template. I need it to sound like a real person wrote it, mention specific problems my customers have, and include a clear reason to reply. Try again.”
Gets: something actually worth sending
Too nice:
“Give me some tips for improving my website.”
Gets: ‘use clear headings, add images, make it mobile-friendly’ — thanks, very helpful
Disappointed parent:
“These are the same tips every blog post gives. I need specific, actionable recommendations for a B2B consulting website targeting mid-size companies. What would actually move the needle?”
Gets: targeted advice you can implement today
Too nice:
“Help me write a bio.”
Gets: ‘[Name] is a passionate professional with X years of experience...’
Disappointed parent:
“This sounds like every LinkedIn bio ever written. I want something that actually sounds like me — casual, confident, maybe a little funny. I’ve been in tech for decades and I’m done pretending to be corporate. Try again.”
Gets: something with personality
Why This Works
AI models are trained to be helpful, which often means they default to safe, broad, inoffensive answers. When you push back, you’re giving it two critical signals:
- More context about what you actually want — “generic” and “too corporate” tell the AI what to avoid.
- Permission to be bolder — most people accept the first try, so the AI plays it safe. When you say “try harder,” it reaches deeper.
It’s not about being mean. It’s about raising the bar.
The Escalation Ladder
- First ask: Get the baseline response.
- “This is too generic” — forces specificity.
- “Can you give me 3 different approaches?” — forces creativity.
- “Which of those would work best for [specific situation]?” — forces judgment.
- “Now combine the best parts into one final version” — forces synthesis.
By step 5, you have something genuinely good. And it took two extra minutes.
The Golden Phrase
If you remember nothing else from this tip, remember this phrase:
“This is good, but I know you can do better. What would the exceptional version look like?”
Use it every time you get a first draft from AI. Every. Single. Time.
One More Thing
This technique works in both directions. If AI gives you something too complex, the disappointed parent voice works there too:
“I asked for a simple explanation and you gave me a textbook chapter. Pretend you’re explaining this to someone smart who doesn’t have your background. Try again.”
The key is specificity in your disappointment. Don’t just say “do better.” Say why it’s not good enough and what better looks like.
Just like you would with your kid’s homework.
Try this on your next AI prompt.
Push once, then push again with specifics. The quality jump is real.
Browse More AI Tips →More AI Tips
Want the rest of the series? Head to the ChuckGPT Blog Index for more practical AI wins.