Idea to Start

I built my first product for myself.

Not to impress anyone.
Not to validate a market.
Just to solve a problem I was living with every day.

That changed everything.

When I build for myself first, I remove most of the hardest early uncertainty in product development.

I already know:

• the problem
• the user
• what “good” looks like
• whether the system actually helps

In practice, that means I become:

• the first user
• the product manager
• the tester
• the feedback loop

And that is a much better place to start than guessing.

For AI systems, this matters even more than in traditional software.

Because the failure modes are subtle.
The outputs can look right while being wrong.
And the gap between a prototype and something that works in production is usually where the real work begins.

My rule is simple:

• start with a pain I personally feel
• use the system myself before I generalize it
• pay attention to where it breaks, not just where it looks impressive
• only then decide what deserves to be built properly

That’s how I reduce noise early and learn faster.

What do you think: do you build for yourself first, or start by designing for a broader audience?