Why these notes exist
A separate place for personal research — AI systems, energy, and the things I am still working out.
Most of what I publish under Archegon has to carry a certain weight. It is project material: sourced, hedged where it should be hedged, and written for people deciding whether to put capital or years into an infrastructure bet.
This is not that. These notes are personal research — closer to a lab notebook than a position paper. Some of it will touch energy and compute because that is what I spend my time on. Some of it will be about agent systems, evaluation, and the unglamorous parts of making software reliable. Some of it will turn out to be wrong.
Why keep them separate
Mixing the two does a disservice to both. Project material needs to be defensible; exploratory writing needs room to be provisional. Putting a half-formed idea next to a resource case makes the resource case look speculative, and makes the half-formed idea look like a commitment.
So: two areas. The research notes stay tied to the Archegon thesis and cite the evidence base. These notes are mine.
What to expect
Short pieces, published when there is something worth saying rather than on a schedule. Code where code is the clearest explanation. Corrections in public when I get something wrong.
If you want to argue with any of it, that is the point — the address is in the footer.