I’ve been thinking a lot about how software engineering is changing in the age of AI agents.
I’ve shared some of these ideas on X and in presentations, but I want a place to develop them more fully — to write longer, cite others, get feedback, and build a body of work over time.
This blog is that place.
What to expect
I’ll write about:
- Agentic development — how to make AI agents build correct software, not just plausible software.
- Spec-driven methodologies — why specifications matter more, not less, when your developer is stochastic.
- Knowledge engineering — building context that agents can actually navigate.
- The future of programming — what changes, what stays the same, and what we should be paying attention to.
How this blog works
This blog is built with Hugo and deployed on GitHub Pages. The source is public — you can see every change in the git history.
Articles are available in multiple languages, translated with the help of LLMs and reviewed by me.
You can leave feedback on any article via the comments section, powered by GitHub Discussions.
Welcome aboard.