How I Accidentally Stepped Off the Algorithm
For the past month or so, I have been on a digital declutter. I only realized this today, because this was not something I planned. I guess it just happened.
For the past month or so, I have been on a digital declutter. I only realized this today, because this was not something I planned. I guess it just happened.
“Vibe coding” is a category error
Why the term “vibe coding” massively misrepresents and trivialises the fundamental transformation of software development as we know it.
Automating coding is not the same as automating engineering. Engineering is not solved, and it's not even close.
So I've posted before how I think ClickHouse is great. Its columnar storage is fast, it has amazing compression. I won't bore you with those details, you can see them all here: https://matduggan.com/clickhouse-is-winning-the-observability-wars/ However all te…
Explore the challenge of predicting the past in web data science, uncovering the effects of survivorship bias and hazard rates.
I found out a few days ago, that regardless of your permission settings on connectors, Claude auto-mode is able to use them. This violated my expectations, and exposed me to the lethal trifeca of AI agents.
An agent deleted infrastructure without asking, the platform's audit log attributed it to a human, and the hook that stops it from happening again.
Claude Code memory usage, measured: three sessions ran 32 processes and MCP servers outweighed Claude in two. A RAM budget for running sessions in parallel.
On 18 August I deleted every comment in a 29-day-old codebase: 71,660 lines across 742 files. Agents wrote most of them, and agents believed all of them. Here is where the reasons live now, what broke, and how the comments came back.
Everything, from the ecological to the psychological to the weird
Why taste matters and how to develop it.
I Built a Stereo Vision Board for Hardware Hackers
This article was written entirely by a human, for humans. I tend to get excited about certain chips. This time, it was a microprocessor: The Rockchip …
My latest vibe coded game experiment is a Command & Conquer style RTS (Real Time Strategy) game called Battle Drones (play it here). Unlike my previous experiments, this one is …
My comparison of ChatGPT Dot, Grok Bot, and Autumn: natural voice, fewer permission prompts, and why my custom Claude Code assistant remains my primary agent.
Learn how Claude's prompt cache works and how to keep your token usage under control.
Developers who loved building things now spend their days supervising AI agents. Nobody asked them. What engineering leaders should be asking instead.
What it actually took to get a driving licence in Germany, why I started reading about flying lessons instead, and the quiet problem with a system where the people selling the lessons also decide when you are ready.
A good redesign requires a plan and careful execution. Otherwise a designer risks breaking established user habits, which, in turn, make users angry.
Software engineer, tinkerer, builder
How I'm changing the way I work to build and preserve human wisdom now that agents handle most of the implementation
Voxtral 3: Mistral's voice agent SDK for real-time speech in 50+ languages — accurate transcription, emotional TTS, deep tool use. EU-hosted or self-deployed.
From Clinkipedia, the free AIncyclopedia
Why we gave Orbeon Forms’ Form Builder an MCP server instead of an embedded AI chat, what it costs us, and where we think this is going.
Recently I found an open source project that met most of my needs. However, there was a missing feature that I needed for my workflow. I looked for signs of life, last commit was nearly a year ago, oof. I was lucky to find someone else's PR for the exact feat…
Working notes on writing GPU kernels with Triton — tiles, program ids, offsets, masking, and the launch grid, with interactive figures.
One of my favorite labs in college so far has been my Digital Logic Design Lab. Two years ago the final few labs of the course involved building a VGA driver. While that might sound intimidating, it was a surprisingly attainable project. (Even if an inordinate…
Feedback doesn’t only come from customers. Every step of your delivery pipeline gives it, and smaller changes make it faster.
A personal reflection on writing, perfectionism, and creative freedom, exploring ideas from Steinbeck, Barthes, Virginia Woolf, and Joan Didion.
Stars measure interest, not usage. After a year of MindRoom, with forks of a Matrix client, two Matrix servers and a Matrix web client, plus 126 workarounds for one AI framework, my rule is: depend on what millions of projects use every day, roll your own when…
Updates and tips about using Large Language Models (LLM) for programming and development
On Saturday, October 17, at 20:06 UTC, Mozilla received a report that private keys associated with eight eMudhra intermediate certificates had been compromised. These keys could allow attackers to issue ...
Matyáš Racek's blog
The Netherlands is building its own Linux OS to break free from Microsoft
The Dutch government is trying to solve its digital sovereignty problem with its own Linux operating system. A local source recently highlighted the development efforts being poured...
A high-performance and high-level purely functional data-parallel array programming language that can execute on the GPU and CPU.
Jim Calabro
DeepInfer: High-performance, large-scale inference framework for DeepSeek models - deepseek-ai/DeepInfer
OpenBSD 8.0
WhatsApp removed hidden notes telling AI not to recommend a regulator, and not to cite GDPR or CCPA once a lawyer came up. We quote both notes.
An interrupt is not a fault.
Some people think that LLMs should not be used in technical fields like programming or mathematics, because those fields are actually art. True or not, this…
Jeremy Scheff poignantly captures the key loss caused by coding agents for career programmers who enjoyed the puzzle of developing code: Programming turned into one of the greatest jobs you …