I have a confession to make.
I built a fully functional Space Defender game, a responsive blog, and deployed it all to the live internet in less t...
For further actions, you may consider blocking this person and/or reporting abuse
The main problem I have with vibe coding is not with LLM's as a tool but with people not willing to learn.
The problem isn't who has written what, but what do you do when an LLM can't provide the solution. It is like living in a house, but needing all sorts of people to fix even the smallest thing. You feel more accomplished once you can fix things yourself.
You call yourself an architect, but you forget an architect needs to have a lot of knowledge to make the decisions.
You call yourself an error whisperer. But how are you going to fix a 500 error when you don't know what part of the logic is to blame? Error messages can steer you in the wrong direction.
IT students don't only learn the syntax, that is more a bootcamp thing, they also learn a lot of abstract knowledge they can use to make code maintainable. And that is big weakness for LLM's, because they are only trained to make code working not to understand mental models.
There are clients that want new things yesterday, but a large part of them are also the people that are not paying. If you want a career from those clients you are going to be in debt most of the time.
There are levels to this. We’re entering a remarkable new era.
Just like traditional software development, there is a spectrum of mastery. Right now, many experienced software engineers are wrestling with the rise of LLMs because they’ve spent years refining their skill, only to see tools emerge that make complex work appear effortless.
In reality, some developers resist LLMs without recognizing that poor results often stem from unclear inputs, weak direction, or a lack of understanding of how to collaborate with the model. The real advantage comes from learning how to work with these systems, how to communicate intent effectively, guide the output, and iteratively refine results.
Mastery in this new landscape isn’t about fighting the tool it’s about learning how to partner with it.
You nailed it right here. Most of the negativity I saw stems from unfair generalized experience.
Bingo!
"Thanks for reading!"
"Thanks for reading!"
I’m a web developer specializing in Python and Django, with experience in both frontend and backend development. I’m currently looking for job opportunities and would appreciate any support or leads. Thanks!
"Thanks for reading!"
hehe vibe coding is really great even tho I am a teenager i still usually use it for my new game resecons. and it helped me a lot turns out if I didn't use vibe coding skills it would take me an about 6 months and 12 phrases of the project while i can take 1 week whole to make a great demo for my game. I really like vibe coding since 2022 - 2024 after chatgpt releases o4 or o3 gpt's.
"Thanks for reading!"
It’s happening.. it’s 2026 already
Still a few days short, though.
"Thanks for reading!"
"Thanks for reading!"
While I understand your excitement about the efficiency of the LLMs (and I am excited too) and the fact that it allows you to do things extremely fast, i can't understand your comparison and the examples.
Basically you are not writing "vibe coding " is "the new era".
You are writing that through LLM you realized you didn't like front end development. You wanted to be an architect instead or an analyst. The LLM allowed you to see beyond that and to understand that your previous chosen profession did not really fit you.
"memorizing textbooks" was never apart of the job, unless - for some unknown reason - someone taught you some really bad habits on how to write code. Also, syntax was almost never a problem for many years now. Efficient IDEs and linting tools exist long before LLMs went public.
And i do understand that its make us feel "smart" when we use "efficient" prompts, but for me "as a senior" it is a no brainer. There is really nothing special to it. Now juniors are proud that "they discovered "precise English command prompting". For me, it is as simple as defining the constraints that i want the AI to apply to the application.
So, yeah, many others don't like LLM because they don't understand prompting and they fail to make AI do what they want.
I adore it. But i "don't like" the fact that it gets away all of creativity and understanding of the code.
While i can create an amazing image or video scene with a prompt, it does not make me an artist. That does not nullify what I made. But a precise sentence in English is not equal or comparable to painting or directing a scene.
But for the observer, it is obvious you didn't create it. In the same way, anyone can create a space invaders game with AI in a "few" hours". That does not make us better coders.
It just makes the space invaders game creation not admirable (compared to the past) anymore.