An interactive WebGL experiment where fluid motion follows your cursor and blends with bold typography in real time.
This piece combines a distortion slider, custom mouse-driven fluid behavior, responsive type layout, and blend-mode compositing to create a cinematic hover experience.
Built with JavaScript, WebGL fluid simulation, and custom UI tuning for smooth motion, scalable spacing, and fullscreen responsiveness.
Interesting.
Maybe all of these effects could be combined into some kind of āStunning Page Templatesā plugin?
Or through some kind of site like Center Flow | React Bits Pro where each ācomponentā has its own page / price
It is actually a really good idea, and Iāve noted it. However, Iām really confused about what is happening overall with the market. Two years ago, if I had created any of the latest sliders Iāve made recently, I would have had a lot of sales. Now it is literally dead. This month, I donāt know if Iāll even make $500.
So Iām at a crossroads. Iāve started working on the course I told you guys about here, and Iām finishing the final project that Iāll be teaching, a really cool radio player. Iāll also mention WPBay in the course. Hopefully, this will help get some traction because itās really depressing seeing what is happening. Imagine being a junior developer now or lsoing your job as a devā¦
I keep thinking about AI, and yes, it is very, very good. But no matter how good it gets, it will never architect your ideas for you, you still need to know what youāre doing, and I donāt think that will ever change, no matter what kind of AI we have. There are just too many details in a commercial project, and that can only truly be handled by the engineer.
Either way, itās a disaster right now. I hope things will get better in the future.
Your shaders are very good so I donāt think it is a āconversion rateā issue.
I also donāt think itās the AI, because AI canāt vibecode what youāre doing.
If I had to guess, probably a discoverability issue. Maybe people used to stumble upon them on Envato but that place is dead now. Maybe try to do some keyword research and then write blogs, posts, including on X, Reddit, etc.
Just throwing out ideas, not trying to sound like a smartass,
Iām also working so much on marketing these days. Iām doing more marketing than anything else: weekly blog posts, bi-weekly youtube videos, paid google ads, paid/sponsored articles. Not enjoyable at all, but it is what it is!
Iām worried about AI, not so much that people are vibecoding their own stuff, but also think of how much competition weāre getting and going to get. Anyone with an internet connection and Claude subscription wants to compete with usā¦
Yes, I think thatās the real issue. Iāll probably go with the free course soon.
About AI ā I tried vibe coding a project, actually the Izubizu slider. It wasnāt even pure vibe coding; I was using it step by step and trying to follow the code. It got me about 90% of the way there, but after that it started messing everything up. I spent one week in the prompt window and about $200, only to abandon the project and rebuild most of it by hand again, the frustration was really bad, it tried to fix one thing only to mess up somethign else again we are doing engenering not slot machineā¦ it was a horrible experience and I will never do that again.
The reason youāre worried is because you see how good it is. But donāt forget ā you have the skill to guide it and actually understand what it does. For a beginner, it will create crap that nobody will really pay for, and this is already starting to show everywhere. I just saw a few videos on YouTube where companies are hiring seniors to fix exactly those last 10% (or less) needed to get the app running.
Another issue with these types of AI-built apps is that they are not scalable at all since the dev has no clue how was made and AI spits function over functions and make spaghety callse every time just try it you will see. Imagine coming back to the app three months later with a model that has no idea whatās already there or worse witha different dev tha has not clue about the app code and logic.. will mess up things you donāt even know where to start fixing. Youāll try to prompt your way through it, but at that point it becomes a gambling game.
So long term, no matter how good AI gets, skill will be more important than ever. Someone still needs to finish that final 10% and turn it into a working commercial app. Without that, all you get is slop that might work at first glance ā until it doesnāt. Then youāll spend an eternity trying to prompt your way out of the mess. That will never work.
The only way agentic coding will work is by following and understanding the code. You can vibe code sections, but you need to make sure you understand those sections and also fix the slop because it will add it every time , you can see in a vibe codded app multiple typs of coding styles is liek the code was written by mutliple devs each with its own style. Otherwise, those last 10% will never get done. And this wonāt change. We are engineers, creators ā thatās something AI still canāt replace.
The more I use it, the more I see this. It gets better and better and seeign how it spits code is impressive, but as I said, without real skill itās just a slot machine.
If you think someone can vibe code some of your plugins without understanding them well, that would be a challenge. Even navigating one of your products and understanding the functionality requires skill, not to mention the code itself. So relax and market the hell out of it ā thatās the game now.
From what I see, all companies are searching for seniors who can babysit AI for now. That confirms everything Iāve said so far.
On the other hand, going back to writing code manually, letter by letter, feels slow as f*ck. So I do love how easy AI makes it to put things together. But you still need to know how to do that. Skill is required, and it will become more and more important because there is a massive dropout happening in tech. There will be fewer and fewer developers out there.
Same exact experience hereā¦
True, but that crap is still your competition, and the quality level is revealed only in time through reviews. And reviews are also quite fake, especially if the seller is smart and good at customer support (or only sells on their own platform and simply fakes it).
So I agree with you but I think we also have to be realistic. Also think of how many people are truly learning with AI as a teacher, It is more accessible than ever before. If someone wanted to learn to do shaders or WP plugins 10 years ago, it would have involved lots of work, dedication, curiosity, browsing through stackoverflow, etc.
Now you can just ask AI to explain it to you like a dummy. You donāt even need to speak English.
So even with everything you said, I still think we get 3-5x or more competition than before.
Yes but they do not go to the grind like we did and this will show on the long therm! Coding is the hardest thing to learn not just to write the sintax but to put everything togheter and this will not change even if AI spits code all over the place is nothign if you donāt know how to architect it, I am finishing now a product and I wrote most of it using agentic coding now I am fixing things when I look at the code literally there are so many style in it is like one hundred devs worked on it, now I have to fix it so that is readable and scalable but I do appreciate it, it saved a ton of time.
About AI products I can guarantee you that nobody will create something even close with any of your plguins unles it just straigt copying the code but even there it will mess it up somewhereā¦
I am not stressed that AI will take my skill away the issue is that the industry is on life support at this point because of it, who know when we will have some sort of recovery because defenitely it will never be the same as it was before AI.
Hope you are right. We shall see how things evolve. I just tried Claude Design yesterday (new product, launched a week ago), I was impressed.
This is the very 1st dashboard design it did, first prompt, first anything:
This is close to production ready. A few iterations and font work and itās there.
We will see, but I try to stay grounded.
