Most dev content is about web frameworks, system design, or career advice. The overlap between “software developer” and “person who writes Csound instruments” or “person who designs parametric OpenSCAD models” is small enough to be distinctive, but large enough to have a real audience.
AI assistance is genuinely interesting in these domains. Not because AI is great at them — it’s often mediocre — but because the gap is revealing. Ask an LLM to write a Csound orchestra for a specific timbral effect and you’ll quickly find where creative intent and domain knowledge matter. That tension is worth writing about.
The projects are inherently tangible. You end up with something you can hear, hold, or demonstrate. That makes for much richer content than “here’s an abstract architecture pattern.”
It attracts curious people. The audience for this kind of blog tends to be makers, artists, musicians, and engineers who cross disciplines — often exactly the kind of reader who engages deeply rather than just skimming for a Stack Overflow answer.
Some other territory that fits your angle naturally:
- Algorithmic art / generative visuals — Processing, p5.js, or even shader code
- Electronics + code — Arduino/Raspberry Pi for actual physical projects, not just LED blink tutorials
- Livecoding and performance — Tools like TidalCycles or Sonic Pi where code is the instrument
- CNC and fabrication — G-code, toolpath generation, the geometry of making things
- Scientific/data instruments — Writing code that connects to the real world through sensors or measurement