50 Weeks of Vibe Coding — What I Built and What Comes Next
After a year of experimenting with AI, I built VibeLearn AI. Now I'm starting a new series exploring how AI can be applied to everyday life.
Over the past year, I have been running a weekly YouTube livestream every Sunday, experimenting with what I call Vibe Coding.
Now, after nearly a year of exploration, it feels like the right moment to move on to the next stage of this experiment.
The series “Fun Vibe Coding”, which I have been running every Sunday, will conclude with its 50th episode on March 8.
When I first started this series, I had no grand plan.
I simply wanted to learn and explore Vibe Coding with AI.
So I made a small commitment to myself:
Let’s just do this for 50 episodes.
And then look back and see what we have learned.
That simple decision turned into a year-long journey.
What emerged after 50 episodes
Over the course of these 50 episodes, something unexpected happened.
While experimenting and studying with AI, I ended up building a system called VibeLearn AI.
Today, I actually use this system for my own learning and content creation.
The idea behind VibeLearn AI is simple.
If I tell AI:
“I want to learn about this topic.”
the system helps guide the entire process.
AI can:
create a learning roadmap
research the topic
organize the knowledge into documents
and even generate YouTube videos if needed
In other words, the workflow becomes:
Learning → Documentation → Content Creation
All supported by AI.
VibeLearn AI GitHub Repository
https://github.com/solkit70/VibeLearn-AI.git
Organizing the past year
Using VibeLearn AI, I also started organizing the ideas and experiments from the past year.
During that process, I redesigned my project website:
Catch Up AI
Catch Up AI Website
On this site you can find:
the projects I am currently working on
Vibe Coding experiments
AI learning notes
activities related to AI4PKM
Looking back, it feels like many of the small experiments from the past year are finally starting to form a clearer direction.
Starting a new series
With the 50th episode approaching, I decided it’s time to begin a new series.
The new series will be called:
AI in Action: Everyday Experiments
In Korean, the title is:
AI를 일상에 적용해 보는 다양한 실험
Instead of focusing purely on learning AI technology itself, this series will explore something slightly different:
How AI can actually be applied in everyday life.
Each episode will be a small experiment.
Pilot Episode: Learning Lawn Care in the Pacific Northwest
Before officially launching the new series, I created a pilot episode.
The topic is surprisingly practical:
Learning how to manage a lawn in the Pacific Northwest (PNW).
Late last year, I moved from the Seattle area to a community called Tehaleh, which is located a bit farther outside the city.
Compared to city living, this area has much larger backyards.
And that created a new challenge.
I had almost no experience maintaining a lawn.
So I thought:
Why not learn this with AI?
Using VibeLearn AI, I asked AI to help me learn how to maintain a lawn in this region.
The system then helped me:
create a learning plan
research relevant information
organize the knowledge into documents
and guide the learning process
You can see that experiment in the pilot episode below.
Pilot Episode 01
https://youtube.com/live/kat7O_Di5qw
What if AI could also create the video?
Anyone who runs a YouTube channel knows that video editing takes a lot of time.
For someone like me—whose channel is mainly about learning and exploration—editing can sometimes take time away from the actual learning.
So I started experimenting with connecting AI-based learning workflows to content creation.
VibeLearn AI already produces structured documents during the learning process.
From there, I can generate videos using a tool provided by the AI4PKM ecosystem.
The tool is called markdown-video skill.
markdown-video skill
https://github.com/jykim/claude-obsidian-skills/tree/main/markdown-video
I extended this skill slightly so that it can generate multilingual videos.
As a result, I can now create:
Korean videos
English videos
relatively easily.
This extension itself was built using Vibe Coding.
A changing culture of software development
Through these experiments, I have started to notice something interesting.
The concept of software development itself may be changing.
Traditionally, building software often meant:
hiring developers
spending months building a product
testing across many environments
maintaining the system over time
But in an AI-driven environment, a different model is emerging.
Someone can build a small tool that expresses a product idea, and then others can:
adapt it
modify it
extend it for their own needs
In this world, the most valuable thing may no longer be the code itself.
Instead, it may be the product idea behind it.
Moving from Vibe Coding to AI in everyday life
For me, this feels like the natural next step.
Until now, I have been learning Vibe Coding itself.
But going forward, I want to explore something different.
How AI can actually be applied in everyday life.
So the “Fun Vibe Coding” series will conclude with its 50th episode on March 8.
After that, I will begin a new series:
AI in Action: Everyday Experiments
About the livestreams
The weekly livestream will continue.
However, there is one important detail.
Most of the livestreams are currently conducted in Korean, which may make it a bit difficult for English-speaking viewers to follow in real time.
Hopefully YouTube will introduce real-time AI translation for livestreams soon. 😉
In the meantime, I plan to continue creating English videos alongside Korean ones, thanks to the AI-based workflow described above.
You can also find both Korean and English documentation in the VibeLearn AI repository.
Future episodes
Future episode plans will be tracked here:
https://docs.google.com/presentation/d/1BPqYMfcEv-XJ8lkg3aTfc8IBzP7D_FJE7Tf9pWtxW1I/edit?usp=sharing
And the videos will appear in this playlist:
AI in Action Playlist
https://youtube.com/playlist?list=PLRQGNaa1hGF0qqKmgIT1nddlhhucTH0WH&si=rDLsEbajcX95mf9Z
This is an open project
All of the projects I run are open projects.
Anyone is welcome to join.
If you have ideas like:
“Could AI be used for this in everyday life?”
please feel free to suggest them.
I would love to try those experiments.
And hopefully, together we can explore many ways to bring AI into everyday life.