Imagine spending several weeks building an application. You've stayed up late fixing bugs, rewritten features that didn't work, changed the design countless times, and finally managed to put together something you're proud of.
You share it online.
Someone asks how you made it. You mention that you used Claude Code, Cursor, or another AI coding tool.
Suddenly, the conversation changes.
Instead of asking what your application does, someone calls it "AI slop." Another person says you aren't a real developer. Someone else warns people not to use your software because it was made with AI.
They might not even have tried it.
This is one of the stranger contradictions surrounding vibe coding. The technology has made creating software accessible to people who might never have written a traditional program, yet the people taking advantage of it can face considerable criticism simply for how they built their projects.
Some of that criticism is justified. Some of it has very little to do with the quality of the software.
And the difference between those two things matters more than people might think.
Why is vibe coding looked down on?
For a long time, becoming a software developer meant learning a fairly specialized set of skills. You studied programming languages, learned how computers interpreted instructions, spent hours debugging problems, and gradually developed an understanding of how complicated systems fit together.
That knowledge took time to acquire. It still does.
Then AI coding tools appeared, and suddenly someone with an idea and an internet connection could produce a working application in an afternoon.
At least, sometimes.
This changes more than the development process. It changes who gets to participate.
Someone who has spent years learning to program might understandably feel frustrated watching a beginner produce something visually impressive by describing it to an AI assistant. The beginner may not understand the underlying code, yet the finished product can resemble something a trained developer would have spent days building.
It's tempting to interpret that frustration as simple gatekeeping. In certain conversations, it probably is. But reducing every objection to professional jealousy would be unfair.
Experienced programmers know how easily software can appear functional while hiding serious problems.
An application might work perfectly during a demonstration and still have broken authentication, exposed database records, insecure dependencies, or code that becomes almost impossible to maintain as the project grows.
AI doesn't make those problems disappear. It can sometimes make them harder for an inexperienced creator to recognize.
That's a legitimate reason to be cautious.
The problem begins when people stop judging the actual software and start treating the method used to create it as sufficient evidence that it's worthless.
There's a difference between criticism and contempt
Suppose someone publishes a vibe-coded application that stores user accounts and private information.
A developer examines it and discovers that users can access each other's records by changing a number in the URL. They explain the vulnerability and recommend fixing it before the application is promoted further.
That's useful criticism. It identifies a real problem.
Now imagine someone sees the same application, notices that its creator used AI, and immediately replies, "Another vibe-coded piece of garbage."
Nothing has been tested. No vulnerability has been demonstrated. The person has simply decided that the development method makes the product unworthy of attention.
Those aren't equivalent responses.
One helps make software better. The other discourages people from participating without telling them what they should improve.
Unfortunately, both can appear in the same conversation, which makes it difficult for beginners to distinguish useful technical advice from hostility.
And because vibe coding is still a relatively new practice, social expectations around it haven't fully settled.
Why people are uncomfortable with the changing definition of a developer
There's an interesting question hidden underneath the arguments about AI-generated code.
What actually makes someone a developer?
Is it writing every line manually? Understanding how the software works? Designing its behavior? Being able to fix it when something goes wrong? Or simply creating something other people can use?
These questions were never quite as straightforward as they might seem.
Professional software development already depends on enormous amounts of work created by other people. Developers use frameworks, libraries, APIs, database systems, visual editors, and software packages they didn't personally write.
Very few people build their applications entirely from scratch.
Of course, using a library isn't identical to asking an AI model to generate an entire application. There are important differences in understanding, predictability, and responsibility.
But both practices demonstrate something fundamental about software development: making software has never required manually creating every component yourself.
AI coding tools push that idea much further.
They shift part of the work away from producing individual lines of code and toward specifying behavior, evaluating results, debugging failures, and deciding what the application should become.
For experienced developers, that can accelerate familiar work. For beginners, it creates an entry point that previously didn't exist.
Neither group can safely ignore what happens beneath the interface.
Still, it's reasonable to expect that our understanding of who can build software will continue changing as the tools improve.
ArtCraft shows why the old assumptions are being challenged
One particularly interesting example is ArtCraft's collection of creative applications.
ArtCraft is developing a collection of free, open-source applications intended to provide alternatives to established creative software. Its projects include PhotoCraft for image editing, VectorCraft for vector illustration, FilmCraft for video editing, and other tools covering photography, motion graphics, document editing, and design.
If those names sound familiar, that's intentional.
ArtCraft is positioning many of these applications as alternatives to the kinds of tools associated with Adobe's Creative Cloud suite.
What's remarkable is the scope of the undertaking.
We're talking about software categories traditionally associated with large engineering teams, substantial development budgets, and years of accumulated technical knowledge.
And the project has attracted attention partly because of its reported use of AI-assisted development.
Its applications are written in Rust, with an emphasis on native performance, local processing, and open-source distribution.
ArtCraft isn't a finished replacement for everything Adobe offers. Several applications remain in early development, and mature professional workflows involve details that can take years to reproduce reliably.
But the fact that these applications exist, can be examined publicly, and have attracted considerable developer interest is significant.
It suggests that the resources needed to attempt ambitious software projects may be changing.
A small team using modern AI-assisted tools can now attempt categories of software that would previously have seemed unrealistic without much larger resources.
That doesn't guarantee the finished result will be competitive. It does mean the cost of trying has changed.
And that is potentially a much bigger story than whether someone considers the development process legitimate.
The interesting part isn't whether ArtCraft beats Adobe
Adobe has decades of experience, deeply established professional workflows, and an enormous ecosystem surrounding its products.
It would be unrealistic to assume that a new project can replace all of that simply by generating enough code.
Yet ArtCraft doesn't need to defeat Adobe commercially to demonstrate something important.
The interesting question is why a project attempting so many complicated creative applications can emerge and attract attention in such a short period.
AI-assisted development appears to be part of that explanation.
What happens when the same tools become more reliable, easier to supervise, and better at maintaining large codebases?
What happens when creators who started experimenting with small applications gain another two or three years of experience?
ArtCraft doesn't answer those questions conclusively, but it gives us a reason to take them seriously.
Why the stigma around vibe coding may eventually fade
Social norms rarely change the moment a new technology becomes useful.
People have expectations about how work should be done, which skills deserve recognition, and what counts as an authentic creative achievement.
When a new tool changes those expectations, resistance is unsurprising.
Consider digital photography.
Editing photographs with software once attracted criticism for making images too easy to manipulate. Today, digital editing is a normal part of professional photography, although arguments about authenticity, excessive manipulation, and disclosure certainly haven't disappeared.
Or consider digital illustration.
Drawing on a tablet doesn't involve the same physical process as painting on canvas. But the difference in tools doesn't automatically determine the quality of the artwork.
The artist still makes decisions about composition, color, expression, and meaning.
Software development is different in important ways, particularly because hidden technical failures can affect security and privacy. Even so, the broader social pattern is worth considering.
New tools can change which parts of a creative process require the most effort. Over time, people may become less concerned with whether someone used a particular tool and more concerned with what they managed to produce.
That seems plausible for AI-assisted software development, too.
If AI-generated code becomes routine in professional workflows, simply knowing that a project involved AI may tell us less and less about its quality.
Instead, more useful questions will become important.
Does the software work reliably? Is it secure? Does it respect its users? Can its creator maintain it? Does anyone actually want to use it?
Those standards aren't going away.
If anything, they may become more important as creating software gets easier.
The stigma probably won't disappear completely
It's worth resisting the temptation to predict a future where everybody happily accepts vibe coding.
Some objections are likely to remain.
People will continue worrying about insecure applications, poor maintenance, job displacement, copyright disputes, and developers publishing code they don't understand.
Those concerns aren't automatically prejudices. Some will require better tools, better education, and more responsible practices.
There will probably also be people who genuinely prefer software written and reviewed through traditional development methods, particularly in environments where reliability is critical.
That's reasonable.
What may disappear is the assumption that using AI makes a project inherently unserious.
As AI-assisted development becomes more ordinary, the label itself may become less interesting.
People might eventually ask what your software does without first asking whether you typed every line yourself.
Why starting now could be an advantage
There's an opportunity in being interested in something before everyone agrees it's respectable.
Not because criticism automatically means you're ahead of your time. Plenty of unpopular ideas are unpopular for good reasons.
The opportunity comes from the fact that AI-assisted development is still changing quickly, and the people experimenting with it today have time to learn how to use it effectively.
Imagine two people who both want to build applications.
One decides not to begin because they're worried about being called a fake developer.
The other starts making small projects, encounters bugs, learns how databases work, figures out authentication, discovers why testing matters, and gradually becomes better at supervising AI-generated code.
After a year, the second person may have several completed experiments, a better understanding of the development process, and experience dealing with real users.
That doesn't mean they've become an expert programmer overnight.
But they've accumulated practical experience that the first person hasn't.
And experience compounds.
Even if AI coding tools improve dramatically, knowing how to recognize a bad implementation, clarify requirements, and understand user behavior will remain useful.
The strongest advantage may not come from being able to generate code quickly. It may come from learning how to turn that code into something people genuinely want.
You don't need to wait for permission to build something
There's a tendency to treat new technology as something that needs cultural approval before it's worth using.
But software development isn't a club that grants permission to create applications.
If you have an idea, you can attempt to build it.
You still have responsibilities. If your app stores personal information, you need to protect it. If people depend on it, you need to take reliability seriously. If you're using AI-generated code you don't understand, you should be cautious about what you release.
None of those responsibilities require you to stop experimenting.
Start small. Make something useful for yourself. Ask other people to try it. Learn why things break and what makes them work again.
If you need help understanding the code, use AI to explain it, but don't assume its explanations are always correct. Read documentation, write tests, and get independent review when the risks justify it.
You can become more technically capable while using tools that make programming easier.
Those ideas aren't contradictory.
The problem with publishing vibe-coded projects online
Building the software is only part of the process.
At some point, you'll probably want to show it to someone.
And that's where the social stigma becomes particularly frustrating.
General programming communities aren't always welcoming to projects created primarily through AI prompts. You might receive useful feedback, but you might also find the discussion turning into an argument about whether vibe coding counts as programming.
Instead of discussing the application, people discuss the method.
For someone who's trying to find their first users, that can be discouraging.
It also makes it harder to get meaningful feedback. Being told that AI coding is bad doesn't explain why a particular feature is confusing or why someone stopped using the application.
Creators need places where the software itself can be examined without every conversation becoming a debate about AI.
Hobbias: A place for vibe-coded projects without the anti-AI hostility
That's part of the idea behind Hobbias.
Hobbias is a platform for sharing free software, including applications, websites, games, and tools built with AI assistance.
Its purpose isn't to decide whether someone deserves to call themselves a programmer.
It's to give creators a place to present what they've built and allow others to discover it, try it, and provide feedback.
Vibe coding isn't treated as something a creator needs to hide.
You can explain which AI tools you used, describe how your application works, and share a link to the finished project without having to pretend you developed everything manually.
And if someone finds a genuine problem with the software, that should still be discussed.
There's an important distinction between creating a welcoming environment for AI-assisted developers and creating an environment where AI-generated software is beyond criticism.
Hobbias aims for the former.
A bug is still a bug. A security vulnerability is still a security vulnerability. An application that doesn't work shouldn't receive praise simply because its creator used AI.
But neither should a functioning application be dismissed automatically because it was built through prompts.
When criticism focuses on the actual software, developers have something useful to learn from.
What can you share on Hobbias?
Hobbias is intended for free software projects across several categories. You might have built a small browser tool, an experimental game, a website, a productivity app, or a utility that makes an everyday task easier.
A project listing gives you space to explain what your software does and how visitors can try it.
If external services or user-supplied API keys can involve costs, those requirements should be disclosed clearly.
The point is to make discovering and testing new software easier, while giving creators somewhere to present their work.
You can explore the available projects and share your own on Hobbias.
What will matter when everyone can build software?
Suppose AI development tools become so capable that almost anyone can create a basic application.
What happens then?
One possibility is that software becomes much more abundant. There are more personal tools, more experiments, more small products, and more people attempting ideas that previously would have required hiring a developer.
That sounds exciting, but it creates a new problem.
If millions of people can create software, merely creating something won't be enough to make it stand out.
People will still need to identify real problems, understand their users, make sensible design decisions, and maintain what they've released.
In other words, making the code easier to produce doesn't make every other part of building a good product easy.
This is another reason to start experimenting now.
The earlier you begin, the more opportunities you have to learn the parts of software creation that don't disappear when the tools improve.
You can learn how people respond to a design, why certain ideas fail, what makes software enjoyable, and how difficult it can be to persuade someone to return after their first visit.
Those lessons are difficult to get from a tutorial alone.
Frequently asked questions
Why do programmers dislike vibe coding?
Some programmers object to the security and maintenance risks of AI-generated code, especially when creators publish software they don't understand. Others question whether generating code through prompts demonstrates the same skills as traditional programming. There is also a degree of cultural resistance to changing development practices, although attitudes vary widely between individuals.
Is vibe coding considered real programming?
That depends on how programming is defined. Vibe coding produces software through instructions given to an AI system, but it may involve less direct code writing than traditional development. More important for users is whether the resulting application works, is safe, and can be maintained.
Will vibe coding become socially acceptable?
It's plausible that AI-assisted development will become increasingly ordinary as the tools improve and more developers incorporate them into their workflows. However, concerns about software quality, security, and responsibility are unlikely to disappear entirely.
Is ArtCraft really an alternative to Adobe?
ArtCraft is developing a collection of free, open-source applications intended to provide alternatives to several established creative tools, including products associated with Adobe. Its projects have attracted substantial attention, but many remain in early development. It should be understood as an emerging alternative rather than a complete replacement for Adobe's professional software ecosystem.
Should I learn programming if I already use AI?
Yes. You don't need to become an expert before starting, but understanding programming fundamentals makes it easier to recognize errors, review generated code, debug problems, and maintain your projects. AI can help with the learning process rather than replacing it.
Where can I publish my vibe-coded project without being judged for using AI?
Hobbias is designed to welcome eligible free software projects created with AI assistance. Its focus is on helping people discover software and provide constructive feedback, rather than dismissing projects simply because AI was involved in their development.
The people who start building now may have an advantage later
A few years from now, the arguments surrounding vibe coding might seem rather strange.
Perhaps AI-assisted development will become so ordinary that nobody considers it remarkable. Perhaps the distinction between writing code yourself and supervising an AI system will remain important, but carry less social judgment.
We don't know exactly how the technology or the culture around it will develop.
What we do know is that people are already using AI tools to build applications, experiment with ambitious ideas, and create software that previously might have been beyond their technical abilities.
Some of those projects will fail. Some will be insecure, poorly designed, or simply unnecessary. A few may become genuinely useful products.
The important distinction won't always be how much code someone typed.
It will be what they managed to create, how responsibly they built it, and whether anybody finds value in using it.
If you're interested in building software, you don't need to wait for the social debate to settle.
Start learning. Make something small. Improve it until another person can use it. Then make something better.
And when you're ready to show people what you've made, share it on Hobbias, a place where AI-assisted development is welcome and the work itself can be part of the conversation.
The tools are changing. You might as well learn what you can do with them.