Practical AI adoption
Telling Your Story
Why Inputs Matter More Than Almost Anything Else
I recently coined a term, with a little help from my friends and a lot of influence from my hobbies: narrative hobbyist.
As someone who builds dioramas, everything starts with context, backstory, and intent. In a single frame, I'm trying to capture a moment that tells a complete story to someone else.
The more I think about it, the more I realize this isn't limited to hobbies. Writers, artists, designers, and war-gamers all do the same thing. They build meaning through context.
Running a business isn't much different.
Every company is telling a story. About who they are, who they serve, and what problem they exist to solve. That story has thousands of subplots, from customer interactions to internal decisions, all supporting the same goal.
AI works the same way.
Without context and intent, it struggles.
With them, it becomes useful very quickly.
That's why inputs matter more than almost anything else.
A lot of attention gets paid to prompts. They're visible. They're easy to tweak. They feel like control.
But prompts are just how you ask.
Inputs are what you're asking about.
A clever prompt can't save a story with no backstory.
When people talk about "AI hallucinating," what's often happening is simpler than it sounds.
There are gaps in the story. And when there are gaps, the storyteller fills them in.
AI does the same thing.
Inputs are everything the AI needs to understand the world it's being asked to help in.
Past customer conversations.
Policies and constraints.
Tone and expectations.
What success looks like.
What must not happen.
Without that background, AI guesses.
Confidently.
With it, AI can draft something that actually fits.
Think about something familiar, like responding to a customer email.
"Write a response to this customer" produces something generic and risky.
Add context about your tone, your policies, the customer's history, and the outcome you care about, and the result changes dramatically.
Same AI.
Different story.
Very different result.
This is where humans still matter most.
Humans decide what background is relevant.
Humans notice when something feels off.
Humans decide whether the story matches reality.
AI can draft the story.
Humans decide whether it makes sense.
This matters even more for small and mid-sized businesses.
SMBs run on institutional knowledge. A lot of what makes the business work lives in people's heads. AI can't help with what isn't shared.
AI doesn't surface what you know.
It only works with what you give it.
So instead of asking, "How do I write a better prompt?" try asking a different question:
What background would a human need to do this well?
If you want better answers, tell a better story.
Once the story is clear, the next question becomes:
Which parts of this story should AI help with, and which parts should stay human?