I think finding customers is harder than building the product.
There are so many small problems that AI can solve today. Most of them are too small for big companies to care about, but that’s exactly what makes them interesting for a solo founder.
The hard part is finding the people who have those problems.
I stopped asking AI to just write code.
Here are 10 rules that changed the way I build software.
1. Goal first. Code second.
Tell the AI what problem you're solving and what outcome you want.
2. Define the MVP.
Ask: “What’s the simplest MVP I could build in one evening?”
This keeps the project from growing too early.
3. Ask for the best option, not every option.
Don’t list every possible solution. Pick the one you’d recommend and explain why.
4. Don’t overengineer.
Don’t ask AI to design the perfect architecture. Ask: “What’s the simplest solution that solves 80% of the problem?”
Start there.
5. Separate facts from assumptions.
Ask: “What do you know for sure? What’s a best practice? What needs to be tested?”
6. Don’t ask for code right away.
Ask: “What should we verify before writing the first line of code?”
Check that first. Then write the code.
7. Ask AI to challenge your idea.
Ask: “Imagine you’re the lead engineer. Why could this project fail? What are the three biggest risks you see?”
8. Ask for the next step.
Ask: “What’s the next step that will make the most progress in the least time?”
9. Ask AI what you’re missing.
“What am I overlooking?”
This often reveals problems you wouldn’t think to ask about.
10. Always end with a next step.
Don’t just give me the answer. Tell me what to do next.
The real skill isn’t getting AI to write more code.
It’s knowing what to ask before the code gets written.
Which rule would you add?
@0_oRai But here's the interesting part: can you really build something great if wealth is the goal? Or does making money eventually change what you're willing to build?