This week is mostly about AI and the reason I don’t want to use it.
This series is still taking shape — I haven’t settled on a fixed posting schedule or article count yet, but I’m starting with five. I collect links in linkding (a self-hosted bookmarking tool), where I add notes as I go, and a Python script turns them into the final post. My goal is to publish every Friday. Comments and feedback are always welcome!
The idea for Open Tabs was inspired by the “Linkdump” series by Dirk Deimeke.
Interesting Articles from the Last 7 Days #
- A great series about the flight booking system and the used operating system TPF. The article The Problem That Built an Industry // a.s describes the system handling 10k transactions per second, but during cheap flight promotions this goes up to 50k transactions per second.
- The essence of this blog post It’s all hallucinations is “Everything an LLM generates is a hallucination, some just might accidentally be true.”
- AI does not summarize, it shortens and so it happens that the meaning gets totally lost. When ChatGPT summarises, it actually does nothing of the kind. – R&A IT Strategy & Architecture does a fantastic analysis on this and show a lot of updates and examples. It is really important to understand, that an LLM does not understand the text.
- A very good summary about AI datacenters and resource waste. The energy and environmental impact of AI and how it undermines democracy - Greenpeace International contains a lot of links to follow and get more insights on the topic.
- In the German talk Tatort 🤡-IT from Graz Linux Days 2026 the presenter was Mika Prokop the Grml project maintainer. I love his stories, check his talks from the previous years too. The lecture is very entertaining, but also very informative.
Background photo by Patrick Tomasso on Unsplash
