I don’t even get what’s the problem with the algorithm anymore. How do people can claim there is a problem if the algorithm is a black box and can’t be measured? Just a gut feeling based on impressions going up / down?

Took a break from vibe coding and vibe marketing with this page turner. So far so good

Tried Atlas from OpenAI. 1. Opened 2. Said "Yes" to a bunch of questionable check marks which I should have probably pay more attention to, If I wanna live when Skynet takes over 3. Run a couple of prompts on one of my projects 4. Closed What am I not getting?

Question of the day. AI influencer is: a) Influencer talking about AI b) AI-generated influencer 🤔

I am frequently using chatGPT 5 to generate images. Have a very specific idea in my head. Very specific prompt.. The result is not always what I am expecting. I try to edit it. And this is how the trivial task to generate 1 specific image launches the loop of endless attempts

Life Pro Tips: If you travel with Deutsche Bahn in Germany your chance that something goes wrong is close to 99%. So if you have a transit on the way or need to be somewhere on time you can already apologize in advance that you are not going to make it.

Touching grass with Dima Pashutskii . Somewhere in Netherlands - Zwolse Bos

Work in progress but no real reason to hide it 😇

My main problem with any social network this day is lack is authentic content. I guess there needs to be a filter or something related to authenticity in algorithm. It can be 2 feeds, one with AI one without. But I just don’t feel good reading the same ideas repurposed with AI

No one probably cares since it has nothing to do with indie hacking. But I just leave it here. Long live Radiohead

"Focus" is my word of 2026. The most underrated word among entrepreneurs, indie hackers and people who want to make their business ideas big. Focus on one thing. Do it good. Grab your rewards.

I spent 10 days of winter in Barcelona and here is what I learned. I like when it is warm.

Interesting day being busy migrating all my activities from ChatGPT to Claude and trying to figure out how to set everything up in a way I stay with the choice.

X Feed is so annoying. Looked once for something related to Claude Cowork use cases. Now my feed is full of people sharing the same kind of AI generated guide in "we are so cooked" manner…

Let’s imagine that in all the AI generated content here I finally found something good on how to use Claude Corowk or Claude code for my use cases. But when I start testing any of these workflows I hit limits pretty fast even before I can understand if this workflow is valuable

Decompressing with this fun weekend read. Heard a lot about red rising and I am challenging myself to shift from classic fiction to sci-fi/fantasy, to see if it helps me relax from busy startup weeks.

I have a spare Mac mini, which is not running openclaw. None of my computers does. Am I missing out?

I guess after watching Chair Company this is my new totem character.

I think this whole Obsidian + Claude thingy is overhyped af. However I connected Claude to my Obsidian. Why? I believe building a personal wiki with AI is actually pretty stupid. The idea of personal wiki is a lot about documenting your thoughts and connecting them together.

How can I push above my limits if today I hit limits on both Claude Code and Linkedin?