TL;DR: I used AI to tighten a 7-minute story into a 5-minute speech—without cutting a single laugh—and won a major storytelling contest. AI didn’t write it, but it helped me rehearse, refine the message, and stick the landing.
A few weeks ago, I entered the NSA Ohio Last Story Standing contest—a five-minute storytelling competition that attracts some of the best speakers out there.
And I won.
But I didn’t win alone.
This is the story of how I teamed up with my AI sidekick—not to write the speech, but to shape it, tighten it, and maybe even sing it.
🎤 The Original Story
Years ago, I competed in the Toastmasters World Championship of Public Speaking with a speech called One Day Younger. It made the top 20 in the world. Seven minutes of story, humor, and heart.
So when the NSA contest came up, I thought, “Let’s bring that story back.”
Except there was a catch.
This contest had a strict five-minute limit, and I’d made a non-negotiable decision: I wasn’t going to cut any of the laugh lines.
Why? Because laughter is the bridge. It’s what makes the story land. And besides—my inner audience member is always hungry for a good laugh.
To keep the laughs, I had to trim everything else down to about 4 minutes and 20 seconds.
That’s a big shift—from a seasoned seven-minute speech to a new format with a stopwatch running.
🧪 Test Speech #1: The Wasp
Before I went with One Day Younger, I tested a different story—one I call Sometimes You Get What You Don’t Deserve, involving a wasp, some pain, and a lot of faux cursing.
I sent that version to ChatGPT, just to run it by an objective eye.
Its response?
“Are you aware there’s a rule about foul language? And you’re getting close. That could impact your judging.”
Yup. AI playing speech cop. (And let’s be honest—getting DQ’d for a fake swear would’ve stung worse than the wasp.)
So I shelved the wasp and moved on to One Day Younger.
✂️ Round One Edits
The first pass was mostly me. I trimmed some chunks from the original version, then practiced and recorded it.
About 20 takes later, I had something I liked.
I sent it in.
Made the top 3.
That’s when things got serious. Because in the final round, the competition was fierce—and two big challenges popped up.
🚨 Challenge One: Cut 40 More Seconds
My own feedback was clear:
“You’ve got great laughs… but if the audience laughs live, you’ll go over time.”
Translation: cut 40 more seconds to leave space for laughs—without hurting the story.
So I asked ChatGPT:
“Give me a bare-bones version of this speech.”
The result? Meh.
Technically accurate. But it read like a user manual for emotions. Helpful only in showing me what not to cut.
Still—sometimes seeing what doesn’t work is just as valuable.
🔄 Challenge Two: The Message Disconnect
Judge feedback:
“Your story is great, but the message doesn’t clearly tie back to your work on stress management.”
Fair.
So I worked with ChatGPT to revise the intro—something that would make the connection clear from the first line, not wait until the wrap-up.
We kicked around a few versions until I landed on one that worked. One that said:
“Yes, this is a funny story—but it’s also a window into how we all manage stress.”
👤 Enter: The AI Judge
Once the new version was locked in, I gave ChatGPT this prompt:
“Pretend you’re a judge in this contest. What would you ding me on?”
It responded. Not with praise. Not with fluff. But with real critique:
-
“Not enough build-up to the emotional payoff.”
-
“Consider making the ‘turning point’ more distinct.”
-
“Double-check your transitions—one is a little bumpy.”
It wasn’t perfect. But it was honest. And instant.
When you’re short on time, having an AI judge who gives you feedback at 2 a.m. and doesn’t mind repeating itself? That’s pretty handy.
🎵 Bonus Round: The Song
With the edits in place, I decided to rehearse differently.
“Turn the speech into a song. Break it into verses like lyrics.”
We weren’t aiming for a Grammy—but we were testing rhythm, pacing, and natural beats.
The speech became a five-part musical:
-
Verse 1: The interruption (“Can we rent jet skis?”)
-
Verse 2: The hesitation (“Dear God, this is a death machine”)
-
Verse 3: The struggle (saltwater, bouncing, 27 MPH terror)
-
The Bridge: The challenge (“Hey old man, wanna race?”)
-
Final Verse: The realization (“I didn’t get older that day. I got younger.”)
Reading it like a song helped me feel the emotional rhythm and spot clunky transitions. It also showed me that Verse 2—the hesitation—needed more tension. Once I added beats to deepen the doubt, the turning point felt more earned.
🎧 The Final Cut
After all of that, I delivered the story live in Columbus.
And… I won.
Not because AI told my story.
Because I did.
But with better timing, cleaner flow, and a sharper message—thanks to a little help from my silicon sidekick.
When you’re competing against Hope and Michael you need all the help you can get
🐓 So What?
I believe in using every tool we have to tell stories that connect.
AI didn’t write for me. It didn’t replace me. But it made me faster, sharper, and a little more fearless.
And hey—when you’re trying to hit 4:20 on the dot without losing a laugh? You take all the help you can get. Just make sure the voice behind the mic is still yours.