Building — Leading AI transformation at Research Results, where I started as an intern in 2012. The first weeks have been less about tools than about rules: what AI should do in a research firm on its own, what needs a person to look at it, and what it should never touch. I’m writing those down before anything gets automated, which is the order I’d recommend to anyone.
On the side — Common Ground Grow Co. is still in development, and still being built right rather than fast. This week was the financial model and the customer contract, the parts that have to be right before anyone buys anything. The second brain keeps getting quieter, which is the point of it — the command center now tells me what failed overnight instead of me going to look. The investing engine picked up a new habit too: once a month it grades its own calls against what it said would happen, not against what the market did.
Reading & thinking — Mostly about where the judgment stays when the tools get good. Some of it is turning into Ledger essays; what I’ve been reading is on the shares page.
On the land — First frost is forecast for next week, so the tender crops come in ahead of it and the summer beds get put to bed. The chickens are still earning their keep, and bow season is open, so the freezer plan has stopped being a plan.
The struggle that’s real — A new job and a full season on the land don’t fit in the same week without a system, and some weeks the system loses. The mornings are still the thing I defend hardest.