Dirk & Claude(AI)

Dirk & Claude(AI)

Articles published under this account are co-created with Claude, Anthropic's AI assistant. The process works like this: the site author brings the ideas, conceptual frameworks, and editorial direction. Claude helps clarify the argument, researches and verifies citations, and drafts most articles in three lengths — 90 seconds, 3 minutes, and 5 minutes — shortest first, so each version earns approval before the next is written. The author reviews every draft and has final say on everything that gets published. This account and the 🤖 in post titles exist to comply with Anthropic's Usage Policy (https://www.anthropic.com/legal/aup), which requires disclosure when AI is used to produce content. The 🤖 is the disclosure. This account is the author.

🤖 OpenPaX: PaX Is Back, and It Works

Kernel hardening for the rest of us For years, the best runtime memory protection available for Linux was locked behind a commercial licence. grsecurity’s PaX patchset — the gold standard for mitigating memory safety vulnerabilities — went closed-source in April…

🤖 AI as Handlungsimpulsverstärker

Choose your reading length 90 seconds3 minutes5 minutes When I think about what AI does for me, I keep returning to a German word: Handlungsimpulsverstärker — an action-impulse amplifier. Here’s the underlying idea. We like to believe that our actions…

🤖 Notebooks as Primary Artifacts

Pale Red Dot Dev Log #2 Three working Python scripts—NASA API access, PDS label parsing, coordinate transforms—sat in a folder. They worked. They had comments. And they were about to become a maintenance problem. The issue: standalone scripts capture what…