Data & AI
Flashing CyanogenMod nightlies in 2009 taught me that understanding the build was never the thing that kept you safe. The nandroid backup was. The vibe coding argument is having the same fight, and mostly aiming at the wrong thing.
You can't prompt Suno AI into behaving the way you want, but you can prompt it into leaving a hole, loop the ten seconds it gets right, or split the stems and mute the rest.
A snapshot of my personal AI dev setup as of mid-2026: Hermes for coordination, Cursor for building, local models on the M4, and the spots where I deliberately skip agents. It'll probably look ridiculous in two months, which is the point of writing it down.
Step one of the 12 steps is admitting you have a problem. So here goes: I deploy agents I don’t need, for tasks a function call would handle, and I live in a constant state of multi-step agent sprawl. OpenClaw on my VPS. Hermes on my telegram. Open models on Ollama, some custom (+ abandoned) ones on my Ubuntu machine. Cursor agents, my daily driver. Claude, also in the melee in a busy terminal. An…
Right now, the reason you got an ad is actually sitting in plain text in a network request that you can read yourself. The browser doesn’t see every server‑side decision, but it shows enough to be informative and useful. I dive deep into personalization and data flows for a living, so I’m in these requests a lot. What I keep coming back to is how much of it you can read from the outside, with…
