Archive
This section contains old Vibecode materials. Some articles are outdated, have been revised or replaced with more current versions, but we decided to keep them in the archive.
Who are you in vibcoding: developer, architect or operator of AI
In vibcoding, you are always responsible for the result. The question is not whether you write code with your hands, but at what level you make decisions
What Humans Are Responsible for, Even If AI Writes Code
Even if the code is written entirely by AI, humans are always responsible for meaning, decisions, and consequences. Code can be generated, but responsibility is not
Why AI bugs are your bugs
AI can write code with an error, but the decision to use this code is made by humans. So the responsibility for bugs always remains with you, even if you have not written a single line
What it means to think architecturally, even without experience
Thinking architecturally is not about knowing terms or patterns. It's about seeing how the parts of the system are connected, why they're needed, and what happens if things change
What you need to know about runtime, even if you don’t write code
Runtime is the environment where the program runs. Understanding runtime helps explain how your program should work and why it behaves the way it does
Why you can’t go without understanding HTTP and API
HTTP and API are a way of communicating between parts of a program. If you don’t understand how a program sends and receives data, you won’t be able to explain to the AI exactly what’s going to happen, and you’ll get code that either doesn’t work or doesn’t work as it should
Why IDE is more important than model
You can take the smartest model and still get stuck. Or you can take a regular model, but in a good IDE, and suddenly start doing things.
Why Starting with a Pet Project is a Mistake
Pet project seems a safe start, but most often it interferes with learning the main thing
How to choose the first task for vibcoding
The first task should not be interesting and not “cool”, but clear and final
Why Copying Other People's Prompts Doesn't Work
An alien prompt doesn't take into account your task, your context, or your level of understanding
How to Read and Verify Code Generated by AI
Code from AI cannot be taken as a ready-made answer
Generation of a large project at once - an architectural trap
When you ask AI to generate a large project at once, you get not a basis, but an illusion of readiness
Why AI almost always makes it “too difficult”
AI almost never complicates out of malice
Lack of logs, tests and checks is the main mistake of the beginner
When there are no logs, tests and checks in the project, it seems that you have made your life easier
Why “It Works” Is Not a Quality Criteria
The phrase “it works” only says that the code is not broken right now
How much time does AI really save
AI saves a lot of time at the start and hardly saves it where there is no understanding
What you should not automate in the beginning
In the beginning, automation often hinders than helps