Fast Product Shipping
In the simplest terms, I can call this an MVP. Nowadays, developing a product is becoming easier, and you don't need to spend months on it, because you need to validate this idea. To do this, you need to explain this product to people and get them to use it somehow. That's why validating the idea by spending less time and budget becomes important at this point.
If we look at where artificial intelligence (AI) has reached today, there are actually many people who design and launch applications within a few days. In fact, since I've been part of the #buildinpublic culture recently, I can easily observe this. I see many people developing applications with vibe coding. As a result, they can build many applications within a few days and test them quickly. They can even reach revenue figures like $10k rapidly. One can't help but be impressed, really.
Since I have more than 5 years of experience as a software developer, I can't fully transition to vibe coding — I can't trust it 100%. I want to have code that is somehow under my control and that I know what I'm writing. But on the other hand, missing out on this process isn't a very logical thing to do. At this point, getting help from artificial intelligence while developing products makes my job incredibly easier. Actually, if I need to explain the path I follow:
- First, I architect the application's core infrastructure by defining the technology stack, selecting appropriate services and APIs, and establishing the system design and architectural patterns to be implemented.
- I develop simple components myself — I take care to make them as reusable as possible.
- For components that will take some of my time, I send the design and the file I'm going to add to ChatGPT and tell it what I want. It then gives me the output.
- The thing I pay attention to here is this; if it suggests something in a complex form, I ask it to simplify as much as possible and make it suitable for my structure.
By doing it this way, I both establish the structure the way I want and develop the project quickly. A component that might take me 4–5 hours if I did it myself, I can handle in 10 minutes with the help of artificial intelligence, and I can test the idea rapidly.
Actually, at this point, rather than saying "artificial intelligence has emerged, so what am I going to do now? I'm going to be unemployed," integrating it into ourselves and our work really saves me a tremendous amount of time. Because I believe that trusting it 100% will make control difficult once the project becomes more complex. That's why I'm convinced that moving forward with small steps is safe.
Best wishes…