Three-dimensional space

He used Jupiter as an example.

For us here in the 3D, it looks like a gas giant, but in a higher dimension, it's fully inhabited.

The civilization is there.

It's more solid.

So if you've been devolved into a 2D reality, then you're gonna experience a flat Earth.

But like in a video game, if there's a flat circle on the screen, just like you're looking at right here, like that moon looks just like a flat disc to us.

But through the evolution of technology, when games went from 2D to 3D, that flat disc becomes a round sphere, right? And then as the technology evolves, it goes from a 3D sphere to a now holographic traversable sphere through virtual reality.

So we can liken our reality to a simulation because we can literally compare it to a video game, and it's the easiest analogy.

And then I remember feeling a little strange and looking around and seeing...

the entirety of what I understood physical reality to be started disintegrate into an overlay of sacred geometry that was dimensionally through whatever we would call 3D physicality.

I had never seen anything like it.

Just that alone, just that awareness and that vision, I see it now, right now, that first moment of it.

Today's episode is part of our ongoing series with Chelsea Weidman, also known as the Geeky Ballerina, where we unpack specific elements of artistry as she identified and explored them in her thesis on artistry and ballet.

In today's conversation, we're digging into what it actually means to move through three-dimensional space.

and why relying so heavily on the mirror can really work against that.

Chelsea also shares an image from her research that will completely change how you think about balance, and we get into ways that we can start conceptualizing and filling the space around us with intention and presence.

So anyway, she has this wonderful quote that I'm going to get wrong, but it's basically, you are not standing underneath the sky.

The sky reaches all the way down to the ground.

And I love it as a poetic idea, but I also think it's a great entry point for three-dimensional space because we never at any point in our training could see the top of our head anyway.

Yeah.

OK, let me try to conceptualize this for you listening, because it is wild once it clicks.

Think of this like a giant 3D coordinate system where every concept in human language is plotted in space.

Yes.

Imagine you have a document in your catalog that says children playing joyfully in the park and another document that says kids happily running around the playground.

A traditional keyword search for the word children would completely miss the second document.

Exactly.

But in this giant 3D coordinate system, because their meaning is basically identical, the mathematical coordinates for those two sentences would be plotted right next to each other.

They live in the exact same neighborhood of meaning.

Para lo último en la carrera de IA espacial, le damos la palabra a Amber.

World Labs de Fei-Fei Li lanzó Atlas el lunes, un modelo único preentrenado desde cero con texto, imágenes, video y 3D.

[carraspea] Puede reconstruir una escena a partir de tan solo dos fotos y generar hasta un minuto de video en 1440p con control de cámara perfecto al píxel.

La imparcialidad de los benchmarks ya está en disputa.

When you want to represent the earth We have to represent it in a model.

So, a map is basically an abstraction of reality which represents the 3D version of the Earth into the two-dimensional.

So you cannot just basically transform from a three dimension to a two dimension without distorting something.

As cartographers they usually use three projections.

So that's the overview of the framework.

So the main contributions for this study is really, you know, the left-hand side is a 3D vision language scene map.

So an integrated vision map.

It's 30p.

We tested 14 fabric types and 33 instructions are tested.

And then it's actually a real-world factory environment validation.

So the first part about planning is a 3D vision language model we are using.

See, we use the deep camera to take pictures around those fabrics that are on the table, and then So here's a parallel processing, right? The left-hand side, we use vision, the images, has the image encoder so that it can, later I'll talk about this, we'll use a 3D cloud reconstruction and then get information of the fabric, like the area, how large, the shape, and the thickness, right? So those physical informations.