2026-02-10 23:16:13
Despite being silent about AI on this blog, I’ve been using it for a while now. And not, it doesn’t write posts here instead of me. But I do use it in programming.
Many software engineers worry that AI is going to replace them. To some extent it might be true, we’ve seen impressive results in the past few years. Many models are now capable of generating basic applications, landing pages, or simple games. But there will be a lot of demand for more complex applications, for critical thinking, and architectural decisions.
However, if we take a look at it from a different perspective, AI also opens up a whole new world of possibilities for everyone. Here’s an example.
I’ve been using a Yubikey for my GPG keys for a long time now.
But I never really liked the GnuPG’s pinentry-mac password dialog used to unlock my key.
It is outdated.
Original pinentry-mac from GnuPG
It became even more apparent on macOS Tahoe with Liquid Glass. But even before that, I couldn’t stand it and quickly hacked my own alternative. I spent a few hours in total to make it work.
My handcrafted alternative from a few years ago
And then some more hours to update it to work with the latest macOS. It worked great and I was happy with it.
But today, I decided to task an AI agent to reimplement it. I started with the following prompt on an empty directory:
Create a macOS native popup that will implement Assuan protocol (at least partially) and act as replacement for
pinentryprogram. It should follow Liquid Glass guidelines and present a simple and clean UI.
And it came up with this while I was brewing a cup of tea. Wow, just wow!
What I got in a few prompts from AI
Well, it didn’t work perfectly from the very first try, but I needed to follow up just twice with a few minor tweaks. Overall, it took me about two minutes of active involvement (and 20 minutes in the background) to get it working.
And AI also implemented the given task in many ways much better than I did. For example, it has better organized I/O and tweaked the UI using properties I’ve never seen before.
We are no longer limited by the current tools. It is now absolutely possible to tailor an application to anyone’s needs. It takes just a few minutes and a handful of AI tokens.
I don’t think the profession of software engineer will ever be obsolete. Instead, AI will enable us to create applications that we wouldn’t be able to make otherwise.
Yes, I spent just a day hacking my own solution for a simple task. But I have many more similar ideas, which I wasn’t able to implement yet because of time constraints and priorities. And now I can!
And I’m not alone. Some are replacing paid services spending the same 20 minutes. Others creating an app that they always needed. These are just two I stumbled upon last week. And I’m sure there are and will be many more.
2026-02-02 02:29:20
It’s been a while since I last posted a changelog for this website. So it’s time to recap what have changed since then.
Two years ago I added a simple XSL transformation to prettify RSS feed as it is shown when opened by a direct link. The support for this (massively outdated) technology is going to be removed soon. Funny, that this specific use case is mentioned in the deprecation announcement. Looks like, I’ll need to add a polyfill to fix that later.
I keep polishing CSS without changing the looks. I’ve cleaned up the narrow view shown on mobile. But the biggest update is the addition of dark theme.
Initially implemented with JavaScript and manually changing classes, now it uses modern light-dark() CSS function and color-scheme property.
Checkout this awesome guide on how to properly use it.
It appears, I like making lists. Not To-Do lists, although them too, but lists of albums I listen, or series I watch. Or my up-to-date wishlist and bucketlist. All that in addition to long existing lists of books I read and videogames I played.
At the end of 2025 I downloaded my data from Instagram and Twitter and vibe coded two static pages with all content I once posted there.
I stumbled upon an article “Why your website should be under 14kB in size” some time ago and decided to implement the advice, at least on my homepage. So I removed all unused and auxiliary stuff, and now it hopefully loads a fraction of a second faster.
Was it necessary? Probably not.
Was it fun to chase it down to 14kb? Hell yeah.
And since I already use a static site generator, why not take another small step to improve load times?
More and more browsers are supporting navigation transitions now, and I’ve enabled them too. With all optimization efforts, pages now switch seamlessly, and I enjoy navigating on my site.
That’s all I could remember I’ve updated here recently. Again, I’m CEO, Product Manager, Head of Design, and Lead Software Engineer for my homepage, so I can do whatever I want here. And I enjoy it!
2026-01-30 16:28:59
Last week I took part in an internal Startup Camp. The goal was to take an idea for a spin for a week, and prove if it is worth further investment or not.
I didn’t have my own idea, so I joined an existing project. We proudly called as co-founders, even in fact we were just learning basics.
I’ve heard it multiple times, but until you start doing something you might be blinded by the absolute awesomeness of your idea. It is only when you start digging deeper you discover all its hidden sides, and may ultimately come to a conclusion that nobody needs it.
It is necessary to believe in the idea, but it is sometimes as well important to let it go when everyone tells you it is not needed, and research confirms it.
During our research, we contacted 50+ people on LinkedIn each. And some of they answered! We did numerous interview just in few days, talking to former employees of competitor companies and potential and current users of similar software. This helped us a lot to understand our problem understanding wasn’t aligned with the reality.
We had an opportunity to continue working on the project if we could have proved it can be successful. Nevertheless, it was also an important experience to close the project before we invested too much into something that can’t be profitable.
They say that only 1 out of 50 startups survives and is profitable. Well, I need to try just 49 more times.
This is ten percent luck
Twenty percent skill
Fifteen percent concentrated power of will
Five percent pleasure
Fifty percent pain
And a hundred percent reason to remember the nameRemember the name by Fort Minor
2026-01-27 23:12:11
After I created my Twitter archive, I started reviewing all my blog posts. The goal was to replace Twitter embeds with pre-rendered tweets from the archive. There were two reasons for that.
First, it ensured the content will remain there even years after publication. For example, my April Fools blog post from 2017 has lost most of its referenced content, because it was deleted sometime in the past years.
Second, Hugo shortcode I used for that is deprecated, and it is required to rename it. I didn’t want to do that. Embedding tweets directly allowed me to work around this and also improve building speed albeit insignificantly.
As a side effect of this activity, I also added some deprecation notes to old content which is irrelevant as of today. I’ve irregularly writing here for around ten years, so there’s been some.
2026-01-16 22:59:00
Just finished reading listening “The Big Five for Life” by John Strelecky (narrated by the author)
and it is one of the most inspiring books I’ve ever read.
Set as a fictional story1 about a successful entrepreneur, it describes his way of leading his multiple companies. But in fact, it goes much further, and focuses on universal life advices that would benefit everyone, not only company owners and leaders.
Here are some easy to remember, but very impactful concepts from the book.
Imagine that at the end of your life, a museum is built that displays every single day you lived. Every action, conversation, and mood is presented there. If you spent 20% of your life at a job you didn’t like, 20% of the museum will be dedicated to it. And you’ll be guiding others thorough your personal exhibition for eternity.
A Good Museum Day is a day you would be proud to be displayed there.
Everyone has their PFE (Purpose for Existing) – the main goal of their life. Yours should align with the company you’re working on. The book focuses on hiring, but this advice may be applied when you’re applying for a work too.
The concept is also useful in private life just as well. If your and your partner’s PFE differ significantly, it may be impossible for you to coexist in balance.
What are the five things you want to do, see, or experience before you die? These are your Big Five for Life.
You should always remember them and ask yourself every day, what have you done to be a step closer to them?
After reading this book, I understood that I don’t exactly what is my PFE, and what items from my bucketlist I would call my Big Five for Life. So, my resolution for 2026 is to figure this out.
★★★★★ highly recommend
Similar to “The Five Dysfunctions of a Team” by Patrick Lencioni. And both have five in their titles. A coincidence? ↩︎
2026-01-06 23:44:10
Few weeks ago I downloaded my complete Instagram profile and published an archive here, on my personal website.
Next logical step was to do the same with Twitter, which was once my favorite social network. I left it right after it was overtaken by a billionaire and renamed, but it hosts around 15 years of memories. I don’t want to lose them, so I downloaded archive of my profile and (vibe-)coded a static page from it.
With very little effort from my side I was able to get a 5Mb HTML file and somewhat 400Mb media.
With some additional effort from AI and imagemagick/ffmpeg I shrinked them down to 2Mb and 120Mb respectively, which I find acceptable.
I enjoyed glancing through the history and remembering some long forgotten interactions with beautiful people around the world. I’ll miss them, but I’m out of social networks.
