This is my week 39 of learning to vibe code without a software development background.

This is my week 39 of learning to vibe code without a software development background. This week I noticed something very interesting. I think the rise of vibe coding is changing the order of how development is done. Let me backtrack. Although I don't have any experience in software development, I have lots of experience in product management. If you're working in tech, you're probably familiar with agile development. You've probably seen that usually it starts with a product vision and strategy, and then the product manager comes in and turns them into PRDs and user stories. Then that gets handed off to a designer who will design the UI and the UX before it goes to the developers to be actually developed, and then finally moves on to the QA team who will do the testing. Now, of course, that is a huge simplification of the process. It's not always sequential like that and people do work together. However, there are points of handoff when you go from person to person, team to team. Now with Solo Vibe coding, you are your own product manager and designer and developer and QA. You are everything, and you don't necessarily need to communicate with anyone to hand off those requirements or vision. So what I noticed is that instead of the entire process like I described, I start where it's most mission critical or where it's most likely to fail. For example, in my holiday card project, the most crucial piece is the image generation. So I spent a lot of time up front really fine tuning that prompt with different styles, different descriptions, changing words to see which version looks better until I was very happy and seeing a consistent output from the prompt. And then I went to build the UI and the UX and the email sending service as well as the backend. Now when I was making this 3D interactive map of the most influential AI experts on X, I kind of knew that the data will be there and the different search functions and filtering functions is a given. But the most mission critical piece for me was the 3D interactive map. It had to be visually appealing and clear, it had to be smart and interactive. What I did was take 30 dummy accounts and prototype that animation and interaction in the middle first, and I fine tuned this a lot until I was very happy with the animation. And then I went and designed the left panel and the filtering function, and then finally towards the end is when I actually got the API to go and pull actual real data and real accounts. Now I started looking back on this because this week I started vibe coding my first ever iOS app, and so I was just chatting with Claude about what I was thinking of, what I needed to do, and asking about different milestones and what I needed to learn. I offered to provide detailed requirements as well as UI/UX designs to Claude, except Claude said no need, you need to get something straight first. Because I am creating a music app, Claude actually said the most mission critical piece is getting the latency down. That is when the user taps or clicks or does something on the screen, they should be able to hear the sound and the output right away. If there's a large latency, this app will have a terrible experience. So before I design anything or refine any requirements, I'm actually going to be testing the different ways where I can get proper sound with low latency on my phone, and then I can add in the UI/UX and the design and refine the requirements. So yeah, that is an interesting trend that I noticed this week. Software and app development, especially for solo vibe coders, don't need to go through the same agile process as what traditionally apps do, and I find it actually better to start with the most mission critical piece.