VSM attractor convergence — empirical results and open questions

I’ve been building a VSM implementation in Rust — five systems, algedonic channel, cryptographic policy veto — and running experiments to see what geometry viability actually has. Over thirty-five runs on physical hardware later, the kernel converges to a stable attractor in ℝ¹²⁸. Measurable, reproducible, confirmed by three independent methods.

The most interesting finding isn’t the convergence. It’s that governance (S5) is redundant at rest and decisive under stress — architecture builds the basin, governance is the difference between surviving a perturbation and not. The experiments make that distinction sharp.

Attaching an executive summary with the main results and some questions I can’t answer from inside the project. If anyone here has worked with VSM in organisational or institutional contexts, I’d be curious what you see.

Luis Pineda R. — luis@tallerpineda.cl