I've been thinking about something lately.
AI has made it incredibly easy to turn an idea into a prototype.
A few prompts.
A little debugging.
And suddenly, there's something you can actually show.
But then comes the interesting part.
Making that thing reliable
That's where the story changes.
You start thinking about:
→ What happens when the user does something unexpected?
→ What happens when an API fails?
→ Who should have access to what?
→ Can this scale beyond the demo?
→ Can another developer understand the code six months from now?
→ And most importantly, do we actually need all this code?
It made me realise something while exploring a few product ideas:
The bottleneck may no longer be building the first version.
The bottleneck is figuring out what deserves to become a real product.
AI can help us move from idea → prototype much faster.
But going from prototype → something people can actually depend on still requires a lot of human thinking.
Maybe that's the next interesting phase of software development.
Not:
“Can AI build it?”
But:
“Should we build it, and what should it become?”
Still figuring this one out myself.
Would love to hear how other developers are thinking about this.
Top comments (0)