My introduction to vibe coding

Hey folks, I've been wanting to learn about vibe coding for a while now. For those of you who don't know what that is, vibe coding is were you interact with an AI and it generates code for you. There are many such platforms and some of them focus on different things like web applications or Python scripting etc. 

The idea is you type or talk to the AI and describe what it is you want and AI generates the code for you and many times it actually runs the code as well. 

There are many things that I want to try using AI that I think can help create a better Blender experience. I'm not going to go into all my ideas at this time but what I want to do is share with you my somewhat brief journey discovering the powers and limitations of a couple different vibe coding platforms .

One of the things I should mention is that I did spend a full day just doing research and creating notes on different vibe coding systems. I knew that I wanted to create web apps, they needed a database, and they need to be able to be hosted as well.

I started taking notes and created this document which is not exhaustive by any means but it was helpful in just getting me started. http://cw1.me/vibe

I also created a PRD, a Products Requirement Document, which was fairly straightforward. I used an AI to help me create it by asking one question at a time until it was completed. I've been asked it to create a markdown document which I can then save to my Google Drive. From my Google Drive I can right click on the document and open it up and Google doc and Google doc will convert it from markdown using my default styles. If you're interested you can see it here:

https://docs.google.com/document/d/1ukQy4oC4olQUW7temecs6zwGfYC0l-FDBlVHeuSf67M/edit?usp=drivesdk

I decided to compare both the Mocha platform and lovable because both of them seem to focus on providing solutions for non-developers. 

I am not a full stack developer and do not know the frameworks that are used like React, Ember, Angular etc. and frankly I'm not interested in learning all that either. I just want to be able to talk to the AI, tell it what I wanted to do and have it proceed. 

So my goal was to create a scaffolding application that I could possibly use in the future. I'm not exactly sure how or what it will end up being but it does have some important components like the ability to use Google authentication to log in, the ability to generate AI images using APIs from replicate.com, The ability to access databases and the ability to host websites. I didn't want to hook up all of these things because I know that can be quite a challenge especially when authentication is involved.

I also wanted to do this inexpensively so just in case it didn't work there wasn't too big of a time or cost investment. And to that end, I did not want to hook up a data store server, like Amazon S3 or other, to store images that were generated. So, I also wanted to enable users to use their own Google Drive to store images if they wish. And of course I needed an admin screen with a specific features for it as well. 

And I wanted to do all of this in 2 days on two platforms. And surprisingly, I was able to complete it, but not without some issues which I'll get them to in the video. 

Please let me know if any of you are interested in this sort of thing and if so I'll keep posting. 

For those of you who want to see the scaffold website and play with it, you can at http://cw1.me/dm .

Anyone that logs in gets a free 100 credits and you can use them to test out the website and generate images etc. The e-commerce component is not yet hooked up. If you need more credits, shoot me a message and I'll get you a hundred more.

EDIT: mocha's CEO reached out to me before this video was posted and gave me a link to help people get started with his platform: https://youtu.be/B87abudwSmc?si=UM6-bK7dHiSD8KPp