The 4090 and the Nano
The same 13 CUDA launches cost 4.78 ms on the RTX 4090 and 39.18 ms on the Orin NX — 8.19×, with residency ruled out as the cause — and the first deploy of the stack to the board found a hang, not a mission.
Superposition / An open notebook
Thinking above the code.
AI is making it easier for me to build. I want to use that opening to understand more: the mathematics behind our tools, the science they make approachable, and the questions worth spending time on.
Notes & investigations
Journal ↗The mark
The psi monogram as geometry — a 738-byte JSON file, an SVG generated from it, and a Blender pipeline whose two runs are byte-identical — with the flat mark and the 3D mesh unable to drift apart.
The ladder
The neuro-symbolic seam as one number — squared Mahalanobis distance from the predictor's mean, computed with the JEPA crate's own routine — plus a rule table, and a healing ladder that is still only a specification.
The agent in the loop
A braid that stores nothing, a write edge invented because two strands are separate processes, a belief clock that lets the map degrade instead of stopping, and an exploration mission that opens itself — with an uncertainty weight of 0.625 measured on the robot.
The front end, rebuilt from lessons
Five existing front ends read as evidence, the lessons written down before the crate existed, then one egui binary with five views, six snapshot tests and a screen that was actually read back — plus a TUI panel and a headless board that can build the GUI but not display it.
Computation is also movement.
A matrix product contains opportunities for reuse. A reduction requires partial answers to meet. What happens when we follow those ideas into the hardware?
Explore the idea in Mage ↗Starting with linear algebra and GPU kernels. Following the connections into music, circuits, materials, medicine, and the wider world as the questions develop.