Sahil's Theory of Everything (WIP)

What is it all?

The short version (so far): it all bootstraps from a single trick, matter that models itself in order to keep existing. That self-model is the seed of information, and eventually of a “Me” that can climb a capability ladder; measure, move and manipulate the world; and live in two worlds at once, the physical and the symbolic. The part that can know itself lives in pattern, in substrate. I’m not building this idea so much as compiling it (slowly, and probably forever), over at unnatural.info. What follows is me trying to walk through it properly, one clumsy paragraph at a time.

Not dying, mostly

Life is about not dying. Everything else, ambition, art, politics, your carefully worded bio, is optional theatrics stacked on top of that one non-negotiable. Even the theatrics tend to serve the rule in the end, they just dress it up enough that it doesn’t sound so grim at a dinner party.

Once you’re conscious of the rule, it stops being a passive fact about your biology and starts being something closer to an act. You don’t just happen to keep existing, you will yourself to keep existing, moment to moment. That’s my version of volo ergo sum, I will, therefore I am, a small quarrel with Descartes. Cogito claims too much. Thinking, looked at closely, is just the manipulation of representations, and plenty of systems push representations around with nobody home, so it doesn’t get you all the way to a self-conscious being in the unnatural sense we actually care about. What does, I think, is willing, and specifically free willing. Wanting something and choosing it, especially against the grain of what’s easiest, is the part that needs a “you” there to be doing it at all.

Modelling a ‘Me’

So how does “not dying” turn into a strategy instead of a coin flip? It starts, I think, with modelling a “Me” good enough to check yourself against, and climbing the capability tree/ladder from there.

In the beginning, life was a matter of (physical) matter mattering to itself enough to somehow continue existing. To do that, logically and physically, it needed a representation of itself against which to measure its continued existence. (Or perhaps it’s the other way round: matter that chanced into that capability was the first real step of life, and everything before it was just reactions, automatic measurements that happened to cause movement, with no “self” involved yet.)

That representation also had to be made of matter, since matter is all that can physically be. But here’s the trick: the matter being represented was more than the matter doing the representing. That extra, the part that doesn’t reduce to the atoms, is what we now call information (thanks, Shannon sensei). Information purposely carried in matter, for some purpose, is what we call a language. DNA is the language of life, binary is the language of the transistor computers we’ve built on top of that, and Thai is the language of the people of Thailand (or was the country named after the language?). Every language represents and talks about some parts of the world, and a few of them, math especially, try to talk about all of it, as objectively as possible.

The unit of a language, matter shaped to hold a piece of information, is a symbol, and the matter of it matters far less than what it’s carrying. Our minds are symbolic. As conscious beings we manipulate symbols with our volition and sometimes against it. In a fractal, chaotic world that never runs short of material to represent, that means we can generate an unlimited supply of (symbolic) Things: ideas, categories, gods, brands, grudges, none of which ever existed as matter at all.

Modelling a “Me” is just the first and most load-bearing of those Things. Once you have one, you’re no longer only reacting to the world, you’re checking yourself against a standard, and that’s the first rung of the ladder. Everything above it is just getting better at maintaining the model, acting on its behalf, and building bigger models on top.

The three Ms

Climbing that ladder seems to run through the same duality of quality every time: does the world (or does “Me”) match what I think it should, or not. That’s a plain yes or no underneath whatever precision gets bolted on top of it later. Out of that duality fall three capabilities, roughly in order of difficulty.

Measurement is checking the model against reality, sensing, comparing, noticing the gap. Movement is acting on that gap by changing your own state or position, still just you, moving through the world. Manipulation is the harder move, acting on the world’s state instead of your own, reaching out and changing something that isn’t you. Each one leans on the last: you can’t usefully move without measuring first, and you can’t usefully manipulate without already being able to move. Most of what looks like intelligence, in a cell, an animal, a company, is really just getting better at this same measure, move, manipulate loop.

Two worlds

All of that climbing happens across two worlds at once, and this is where most of my confusion tends to live. There’s the universe, the physical world, matter and energy doing what matter and energy do, indifferent to whether anyone’s modelling it or not. And there’s the world, the symbolic one, built from the languages and symbols in the section above, meaning laden, argued over, agreed on mostly by convention rather than physics.

We’re symbolic beings stuck in physical bodies, which is a strange arrangement once you sit with it. It means even purely made-up symbols, ones with no direct physical referent at all, end up with real power over physical reality. Money is paper and belief, a border is a line nobody can see, and a promise is a sound wave that outlives the sound. None of it is physically necessary, and all of it moves physical bodies around anyway. That’s the whole reason we bother with language, spoken or written, we’re using it to stay on the same page with each other (as much as possible), since none of us has direct access to anyone else’s “Me,” only to the symbols they hand us.

Reality and perception

Which raises the obvious problem: if everyone’s running their own symbolic model, whose model is right?

The universe doesn’t need anyone’s model to be true, it just is. That’s reality, objective, singular, indifferent to opinion. But nobody has direct access to it. What each of us actually gets is perception, a subjective read, filtered first through senses and then through whatever symbols we happened to inherit or invent. Objectivity is the (probably unreachable) limit where enough perceptions line up closely enough that people stop arguing, math being the language that tries hardest to get there. Subjectivity is everything short of that limit, which is most of everyday life, most of the time.

I don’t think that makes reality unknowable, just expensive to know, and permanently approximate. Every “Me” from earlier is a subjective guess at an objective situation, checked and rechecked by measurement, movement and manipulation, never quite finished.

Which is roughly where the scaffolding runs out and the actual work continues. I keep chipping away at the fuller version of all this, slowly, over at unnatural.info. This page just holds the bones.