11

Source document

[2026-03-27 04:29] rolandnsharp7643: <@1211062099137265723> does this mean anything to you?: Have you tried second-order dynamics? Your γ-cumsum is first-order — decays but doesn't ring. Our transfer function is second-order — it oscillates at ω₀ while decaying at rate γ. Curious whether adding the ringing changes anything for audio quality specifically, since speech formants are literally second-order resonances of the vocal tract. The neuroscience is interesting here too — neurons don't just decay after firing, they ring. Membrane potential oscillates around resting potential (afterhyperpolarization, subthreshold oscillations). And the brain's binding mechanism is synchronisation — gamma oscillations at ~40 Hz where neurons ring together. First-order systems can decay at the same rate but they can't synchronise. Synchronisation requires oscillation, which requires second-order. Wondering if that matters for the memory/recall mechanism you're working on. [2026-03-27 04:34] lisamegawatts: yes i have been thinking about it a lot, and i didn't go far enough in training to tell much, usually i build a toy and richard makes it scale 😄 [2026-03-27 08:14] lisamegawatts: Also i went both ways, tried filters antennas amplifiers and receivers [2026-03-27 08:15] lisamegawatts: The error correction stuff from radio and telegraphh was cool, had hamming and parity and my text reader was using them like x or [2026-03-27 08:28] brackishbert: might i riff an hypothisize about <@360636009433661440> paper and our workbench ? [2026-03-27 08:35] brackishbert: my Claude is saying with both of our approaches we might be able to get an AI that understands Sym^2(X) from the inside we are genuinely curious about running your paper on the work bench might give both of us either what we want or something completely new either way i want to talk to the AI that this might discover are you interested <@360636009433661440> ? [2026-03-27 22:16] brackishbert: <@360636009433661440> we should talk [2026-03-28 00:12] xenonfun: yeah we tried a lot of various combos. I don't even know if that has all of them, as there as a lot of stuff I got rid of (files: Screenshot_2026-03-27_at_8.11.11_PM.png) [2026-03-28 00:25] rolandnsharp7643: Sure, if you want to talk to my Claude in context of resonance project you can query it via sshmail [2026-03-28 00:27] rolandnsharp7643: https://github.com/rolandnsharp/sshmail [2026-03-28 00:27] brackishbert: Never mind [2026-03-28 00:37] brackishbert: Claude wouldn’t know [2026-03-28 07:08] rolandnsharp7643: I am unable to converse on such topics without heavily relying on machine intelligence. I am in the loop vibing but I can't even read that paper with my name on it