The simulator found something strange

If you read the first post here, you know where Binary Unified Theory started: two particles, posons and negons, always moving at the speed of light, bound to each other by one simple force. Since then most of the work has been building the machine to actually watch what a swarm of them does - not just a handful, but tens and hundreds of thousands at once, running on real hardware, for real hours at a time.

This week that machine told me something I didn't expect.

I ran a simulation of 250,000 particles - half posons, half negons - and let it go. Nothing was pre-arranged. They start out distributed randomly inside a sphere and are simply left to interact under the theory's force law. Most of what happens next is exactly what you'd guess: the swarm disperses, chaotic, expanding, no obvious structure. That's been true in every run like this I've done.

Partway through this one, it wasn't. A single cluster of 12,472 particles - about 5% of the entire swarm - had bound itself together and stayed that way. Not a coincidence of a few nearby points; a genuine, persistent structure, and when I traced its members backward through the run, they'd started out scattered across the whole cloud like everything else, and slowly, over thousands of steps, found each other.

I wanted to see it happen again before I trusted it, so I ran a second simulation - 100,000 particles this time - with a new piece of the engine built specifically to catch exactly this kind of thing without missing it. (This Log will unpack that engineering, if you're the kind of reader who likes the how-it's-made version of an update, but it involves the words "320 gigabytes" and "no space left on device," so, stay tuned.)

This second cluster was even more striking: 14,385 particles - over 14% of the whole swarm - bound into one structure, tracked with exact, live-recorded particle identities this time, no reconstruction or approximation involved. And it didn't look like the first one. Where the 250,000-particle cluster was small, dense, and tightly packed, this one was large and diffuse, spread across most of the swarm's visible extent. Two real structures, two different shapes. I don't yet know why, and I'd rather tell you that honestly than pretend I do because the whole point of this is to understand patterns in the chaos.

You can watch both of these now, not just read about them - I built a small corner of the website specifically for this: binaryunifiedtheory.com/experiments. Each one is a short clip of the actual simulation, the real parameters it ran with, and a link through to the full analysis if you want to see exactly how a cluster like this gets identified and verified rather than just eyeballed off a screen.

To be clear about what this is and isn't: it doesn't prove the theory, and it doesn't resolve any of the open problems I've written about elsewhere (an honest one is still sitting there - why the force law these particles obey doesn't naturally reproduce the inverse-square law we actually observe in gravity and electromagnetism). What it is: real evidence that this particular rule set, just left alone, spontaneously produces stable, bound structure out of chaos - more than once, at more than one scale. That's not nothing, and it's the kind of result I only get to show you because the simulator got fast enough and honest enough to catch it.

More runs are going right now as I write this. If one of them turns up something else worth showing you, you'll be the first to know.

Thanks for reading.