Your metaverse is too small

It might go without saying, but as to today, right now, there is no such thing as the metaverse. What exists today are a few 3D experiences and games which serve as walled gardens whose proprietors have a vested interest in trying to own the collective notion of 'metaverse' such that they outcompete the landscape into a monoculture. An actual metaverse, in the sense I believe most people mean is a work of pure speculative fiction about the future of how we as a species communicate and interact, and it's a work that I think has been largely taken at its word by the business world due to one of the primary drivers of our current economic climate: FOMO.

What is it exactly people mean when they use the term? Well, this ranges between everything from some vague hand-waving about NFTs to second life. However, I would posit that what the majority of people mean when they say 'metaverse' is an interconnected web of spatial content addressable and navigable in a manner similar to the way we navigate the 2D web today. A web 3.0 you might say (minus the blockchain). This definition will suit our purposes well enough for the discussion I'd like to have.

The issue with viewing the metaverse as merely spatial is that it becomes limited to just three dimensions, which in my estimation is far too few. If we assign each point in space in the metaverse a vector (x, y, z) and then add up the number of points, perhaps taking our universe as a starting point we end up at roughly 4 x 10^80 cubic meters, or the size of the observable universe.?

However, what if we take the size of a small image say 256 x 256 pixels and treat each pixel in that image as a vector? Even in a simple color space (hex), for each pixel in the image there are 256^3 possible color values. The number we arrive at is about 1.19638... × 10 ^ 40403562.?

The size of the universe is beyond our comprehension. This number lacks a way of describing how incredibly vast it is. Within this number is every image that currently exists, every image that will every exist, and every image that will never exist. And this is just the tiny resolution of 256 x 256. Think about higher resolutions, think about video, think about audio content, think about the exponentially?boundless realm of content lying there waiting to be explored.

What machine learning models are allowing us to do is to begin to navigate and explore this latent space. We're able to generate images and digital content on a scale previously undreamt of, and through this generation we uncover more and more of this boundless space. As we move closer and closer toward simply conjuring up any thought as image, we approach the very notion of human imagination itself made manifest; anything that we could dream, can think of, can create, and will ever imagine, all addressable, navigable, and there for us to simply will into being.

The idea of an interconnected web of spatial content is really quite exciting and promises to reveal entirely new modes of communication and thought, but perhaps we're looking at a few too few dimensions. Why stop at three?