Anyone who frequents tech-related forums has doubtless seen some posts by now where people advertise something cool they’ve built with AI and then get crucified in the comments by keyboard warriors standing up for intellectual property, the dignity of human labor, preservation of the environment, etc. I am on the side of the vibe-coders and find negative comments of this kind quite ridiculous.
Yes, AI has created many problems for humanity and the environment. However, the vibe-coders are not contributing to these problems in a meaningful way; if there were a large-scale movement to boycott the use of LLMs, the backlash against vibe-coders might be understandable, but amidst a sea of AI users, these guys shouldn’t be villainized for using a tool almost everyone else is also using. If anything, the hate towards vibe-coding is counter-productive. Many vibe-coders are creating amazing free open-source software and trying to extract some positive value from AI despite all the negative impacts it’s having on the world. The anger towards vibe-coders is misdirected and discourages the few people who are trying to make the best of a bad situation rather than spending all their time unproductively griping on Reddit about large corporations.
This Reddit post and the comments on it are what got me irritated enough to finally write this blog post. Someone made a cool pixel-art desktop in Rust using Claude – half the comments are appreciative of this person making something cool and sharing it with the world for free, while the other half are mad at him for having used AI. Let’s review some of these comments to go through people’s main arguments against AI in hobbyist programming:
All slop must die.
Calling this ‘slop’ is a bit inaccurate because this project is actually quite well-made. It looks sleek, runs with low resource usage (multiple other commenters have confirmed this with exact numbers), and has detailed documentation. Moving on to the next criticism…
Long term maintenance. Someone who has a bright idea and generates functional slop with a few prompts still doesn’t have the slightest clue how to update and support a project in the indefinite future.
It’s very simple:
Software engineers using AI = good
Random lay people using AI to create slop = not good
The problem here is that this commenter assumes the original creator is under an obligation to “update and support a project in the future.” The project’s creator has made this as a small personal project for himself, and whether he wants to maintain it for a large community, micro-optimize, and fix minor bugs is entirely his choice. If this commenter is really so interested in long-term maintenance, he should start contributing to the Github repository which the project’s creator has very graciously made public. The whole point of open-source software is that someone can make a project without having to bear the burden of long-term maintenance all on their own and instead lean on a community of people passionate about the project – criticizing vibe-coders for “oh but they can’t maintain it long-term on their own” ignores the very ethos and motivations behind open-source.
Also, even if there isn’t long-term maintenance and micro-optimizing, why is that so bad? This is a hobby project, not a core part of the Linux kernel – people have a right to build something for fun without the intention of maintaining it for years into the future. If someone makes something for kicks and is happy with the MVP, more power to them.
The biggest problem with this comment is the very philosophy behind “random lay people using AI to create slop = not good.” AI has empowered coding-illiterate people to build software custom-tailored to their needs and preferences without investing hundreds of hours to learn skills that they don’t regularly need. If someone wants to use Linux for the privacy, customization, etc but couldn’t previously get deep into ricing (customization of visual interface) because of unfamiliarity with programming and not enough time/interest to learn, we should be happy that AI has made this very enjoyable hobby accessible to more people, and that people previously not engaged in ricing are now able to make tools that veteran ricers are also using because they’re so impressed by the AI-generated program. More joy in the world is good, more useful tools in the world is good. It boggles the mind that some people are mad about these positive developments because the “wrong people” are making these tools.
It’s valid to criticize vibe-coded projects for poor performance, bugs, or other problems related to what the AI has actually produced. It is similarly valid to criticize vibe-coders if they misrepresent the way they build their projects and try to pass themselves off as more talented programmers than they actually are. However, screaming “AI slop must die” is childish and makes little sense. I hope that as time goes on, more of us can learn to appreciate the cool projects being built with AI and help with long-term maintenance (something many vibe-coders struggle with), rather than unnecessarily spewing hate online at people who are just trying to help others.