ISSUE NO. 01
Think With AI,
Not At AI.
You do not need the perfect prompt.
Start with the unfinished thought. Speak it. Explore it. React. Narrow. Challenge. Compress.
QUICK VERSION
READ THIS IF YOU ONLY HAVE TWO MINUTES
The conversation
is the process.
The most useful AI conversations I have are rarely built from perfect prompts.
They often start with me saying something like:
I have the start of an idea,
but I do not fully
understand it yet.
That is enough.
The ONE Credit system started exactly like that.
I began with a rough thought about the future of ONE Academy.
The conversation moved through several versions before we reached the stronger idea.
The method is simple:
Speak the rough thought.
Do not wait until the idea is polished.
When possible, use voice-to-text.
Talking usually gives the AI more context than typing a neat sentence.
Ask it to expand.
Useful questions include:
- What am I missing?
- What could this become?
- What does this unlock?
- What is weak about this?
React honestly.
Do not accept the answer because it sounds confident.
Say:
- That part is good.
- That part is wrong.
- Forget that section.
- This gave me another idea.
- Go deeper on this one point.
Narrow.
Follow the strongest idea.
Remove everything that is not helping.
Stress test.
Ask AI to attack the idea.
What would make it fail?
What assumption are we making?
What would a skeptical customer think?
Compress.
When the conversation has become large, ask:
What have we actually concluded?
Then save the conclusion.
The first answer is not the product.
The conversation
is the process.
THE WHOLE IDEA
SPEAK → EXPLORE → REACT → NARROW → CHALLENGE → COMPRESS
You do not need the perfect prompt.
GO DEEPER
Stop trying to write
the perfect input.
Most people are taught to use AI as if the goal is to produce the perfect input.
They search for prompt formulas.
They copy templates.
They worry that they are not asking the question correctly.
That can make AI feel more complicated than it is.
The most useful change I made was almost the opposite.
I stopped trying to sound clever.
I started giving AI the real thought.
A real
ONE example.
The ONE Credit idea did not begin as ONE Credit.
It began with a rough thought about the future Academy.
The first idea had two competing versions.
Maybe the future platform would be subscription-based and EDGE owners would get access included.
Maybe an annual subscription could somehow include EDGE.
Neither version was fully formed.
The important thing was that the conversation started.
The AI expanded both directions.
I reacted.
One suggestion triggered another thought.
Then another.
Eventually I noticed that the bigger opportunity was not the subscription structure at all.
It was making the Academy valuable enough that people would want to stay.
That changed the direction of the conversation.
Later, when store credit came up, the AI initially framed it as a way to pull customers toward physical purchases.
That was not the strongest idea.
I realized the credit itself needed more places to be useful.
That human reaction produced the better model.
This is why disagreement is valuable.
Use voice when
you are thinking.
There is a large difference between typing to AI and talking to AI.
Typing is excellent when precision matters.
Voice is excellent when discovery matters.
When I type, I naturally edit the thought before I send it.
When I speak, I am more likely to say:
- What happened before.
- Why something feels wrong.
- What I am worried about.
- The contradiction I cannot resolve.
- A new idea while I am still explaining the old one.
That is useful context.
Voice-to-text gives you the richness of talking without losing the advantages of a written record.
You can search it later.
You can turn the conversation into a document.
You can ask the AI to extract the decisions.
You can update a project from it.
Talk first.
Polish later.
For idea development, I would teach people to talk first and polish later.
AI as a
possibility engine.
AI is good at generating nearby possibilities very quickly.
It can say:
If this is true, perhaps this follows.
If you remove this assumption, another option appears.
If the customer sees it this way, the economics change.
That is useful, but it does not mean every suggestion is good.
Your job is to notice which suggestion matters.
The human brings
significance.
At ONE, a technically possible idea is not enough.
The idea must fit what ONE is trying to build.
A suggestion may be commercially clever but violate Skater First.
It may create revenue but weaken creator economics.
It may be interesting but add unnecessary complexity.
AI can generate the possibility.
I decide whether the possibility deserves to survive.
AI generates possibility.
The human decides
what matters.
How this translates
outside skating.
You do not need to run a skating company.
Imagine you are deciding whether to change jobs.
Do not ask only:
Should I take this job?
Talk through:
- What you like about your current job.
- Why you are considering leaving.
- What the new role gives you.
- What it takes away.
- What you want your life to look like in two years.
- What you are afraid of getting wrong.
Then react to the AI's reasoning.
The conversation becomes much more useful because the AI is reasoning from your actual situation.
A simple
working loop.
Use this repeatedly:
That is more important than memorizing clever prompts.
The goal is not to make AI think instead of you.
The goal is to make
your own thinking
travel further.
THINK WITH AI · ISSUE NO. 01
Speak.
Explore.
React.
Think further.
The goal is not to make AI think instead of you. The goal is to make your own thinking travel further.
ISSUE NO. 02