5K Followers, 500K Views, and 1K GitHub Stars

Want to document this historic moment.

For the past few months I’ve been posting actively on Instagram, and here’s the result.

I already have five thousand followers, and my most popular video hit 500 thousand views.

Oh, and I also have a thousand stars on GitHub.

The money side is still pretty modest, but I have a feeling 2026 is gonna be absolutely fire.

Lopaka.app Now Supports LVGL — The Biggest Update Yet

Hey everyone, I built a graphics editor for microcontrollers — Lopaka.app (already told you about it). And this time I have, quite possibly, the biggest update I’ve been working toward for over a year:

I added support for the LVGL library. This is the most popular library for building interfaces on microcontrollers. And now it has a proper editor.

For now it’s just basic functionality and some core elements: text, panel, button, slider, switch, and checkbox.

You can change colors, sizes, border radius, and of course the source code is generated instantly.

I hesitated for a long time about whether to build yet another editor, but realized that what LVGL Editor does is too far from my familiar Figma workflow. And SquareLine’s editor is just too convoluted — I find it genuinely painful to use.

Now Lopaka.app is the only microcontroller graphics editor in the world that supports all the popular libraries!

So check it out, let me know what you think, what could be added, what could be improved. And definitely tell your friends about this. Thanks for reading.

AI Doesn't Kill Artists — It Kills Those Who Stopped Being Artists

Everyone keeps saying that artificial intelligence is killing artists, killing creators, and all that

But I’m more drawn to the opinion that AI kills those who stopped being creators.

This is a very powerful catalyst for change. Like in any crisis: some adapt and use it to leap forward, while others, unfortunately, don’t.

Some slide down and get stuck in the same old routine, and it grinds them up.

A huge wave is coming. And we have a choice — to become a master who knows how to work with this wave, to make it swirl around you.

That is the cutting edge of art right now.

I don’t have a drop of fear for the new things coming into our lives.

When chaos breaks through, it increasingly shows its true face through the cracks in old structures that past generations imposed on us. Adapting to new technologies is understanding the “algorithms” of how consciousness plays.

Consciousness dwells in bliss within its own wildness and desire to resist everything. And we are the epicenter of these changes. In old clothes, in old “spacesuits,” we won’t master these new spaces. We need to create new identities. ©

The Discovery Problem for Vibe-Coded Apps

Actually, I’d love to be able to find an app from ready-made ones, but unfortunately Google sucks at this. Searching GitHub can also take half an hour. Plus you need to download something, install it, compile it, run it, and figure out what the author was trying to say.

But there’s no way to search like this. And it seems fundamentally impossible. I thought about making some kind of catalog for vibe-coded apps, but I imagine search would be just as big a problem there.

Building Is Now Faster Than Searching

We’ve reached a point where building your own app for a specific task is way faster than finding such an app or service on the internet.

I’ve been through this myself several times already, and again I was trying to figure out how to use the new video generation model from XAI. And everything they have ready-made is super limited and janky, and doesn’t solve the task I actually need.

In the end I wrote a React app in half an hour, hooked up an X.ai key there, and did everything I needed (the model turned out to be meh).

And besides finding something, you still need to choose from thousands of similar apps, figure out if they have the features you need, go through onboarding, pay for a subscription.

Plus somehow you need to transfer context. That is, what I want to do and what data or constraints or requirements I have, trying to fit them into the framework of a ready-made solution.

So if you can formulate a task in a few voice messages, and this task repeats more than once, it’s now much easier and more profitable to build your own than to search.

Code is super cheap now, but attention keeps getting more and more expensive.