Skip to main content
Two-dimensional space

Two-dimensional space

Search complete. 66 mentions across 14 episodes found for "Two-dimensional space".

Sep 12, 2026

John MoustakasGUEST
1:50
Uh, keeps it interesting.
Rob SparksHOST
1:52
We're gonna discuss the DESI Legacy Imaging Surveys today, the largest 2D map the universe ever created.
Rob SparksHOST
1:57
First, why were these under-- surveys undertaken, and what were their goals?
John MoustakasGUEST
2:01
Yeah.
speaker_1HOST
22:44
Exactly.
speaker_1HOST
22:45
To solve this geometry problem natively, they implemented 2D rotary position embeddings.
speaker_1HOST
22:50
We talked about rope earlier for text, keeping words in order.
speaker_1HOST
22:53
The 2D version is a mathematical function that intrinsically captures the spatial geometry of the image.
speaker_1HOST
22:59
It tags every single patch, not just with its visual contents, but with its exact X and Y coordinates relative to all the other patches.
speaker_0HOST
23:07
That is brilliant.
speaker_1HOST
23:08
The language model suddenly understands the image's two-dimensional structure natively without needing a secondary interpolation model to explain it.
speaker_0HOST
23:17
It turns a 2D image into a sequence of mathematically tagged squares that the language brain can just digest naturally.
Wesley EssinkGUEST
11:19
Now, Image data, there's many different ways that we can use image data.
Wesley EssinkGUEST
11:25
And the bit that's always quite tricky is that, obviously, an image data is 2D.
Wesley EssinkGUEST
11:29
And we want to take that 2D data and map it to something that's more 3D, like a surface.
Wesley EssinkGUEST
11:34
So we had a tool called WorkMap, where we can take a sort of transformation from a planar surface to a 3D surface.
Wesley EssinkGUEST
11:44
So in this case, we can take an image that had the text warm map, and we can use it to map it onto a specific area of the model.
Wesley EssinkGUEST
11:52
And I can edit that mapping so I can handcraft the positions and make sure that it fits really nicely onto the surface.
Wesley EssinkGUEST
11:59
So I can use that 2D data on the 3D surface to either emboss or engrave.
Wesley EssinkGUEST
12:05
Or we can use it to add some texturing.
Darren McKeeHOST
35:32
And it's very hard because your eyes are really good in many ways.
Darren McKeeHOST
35:35
And of course, the way your eyes can process different depths, it's very hard for any static image, 2D image to represent that back to you.
Darren McKeeHOST
35:40
That's why things don't, that's why photos look different than real life, duh, right? All I have to say is that because I know the cameras are getting better, it is plausible, maybe not now, but near future, two years, five years, 10 years, where you could get a camera that zooms in on the moon without AI enhancement that is as good as my DSLR from like 12 years ago.
Darren McKeeHOST
35:58
And then you're taking, oh, that's a quote unquote real photo of the moon.
James HarlandGUEST
26:23
Again, it's about collaboration with people saying, I've got a problem.
James HarlandGUEST
26:27
I might have engineers who have to be able to visualize a 3D object from a 2D drawing.
James HarlandGUEST
26:33
There's a lot of 2D drawings in engineering where you say, here's how this works.
James HarlandGUEST
26:37
But there's a skill involved in taking that 2D drawing and imagining the 3D orientation of it all.
James HarlandGUEST
26:44
Well, one simple thing a colleague of ours did years ago was say, well, here's a 2D drawing.
James HarlandGUEST
26:49
Here's a 3D object based on that 2D drawing.
James HarlandGUEST
26:52
Right.
James HarlandGUEST
26:52
So you can test or teach people about, oh, that's what this orientation is.
speaker_4HOST
6:12
Right.
speaker_3HOST
6:13
And what Maldacena's math proved is that these two radically different systems, the 3D gravity-filled bulk on the inside and the 2D gravity-less boundary on the outside, are mathematically equivalent.
speaker_4HOST
6:25
So everything happening inside that warped 3D universe, even the gravitational pull between objects, is perfectly encoded in the complex quantum entanglement happening between the particles sitting flat on the 2D boundary.
speaker_3HOST
6:38
Yes.
speaker_4HOST
6:39
The analogy I always go back to, because this stuff is so hard to picture, is to think of the universe as a solid soup can.
speaker_4HOST
6:55
Basically all the messy physics we associate with moving through space.
speaker_3HOST
6:58
Right.
speaker_4HOST
6:59
But the flat 2D paper label wrapped around the outside of the can has no gravity at all, yet everything happening to the soup inside is perfectly encoded in the ink on that flat label.
The Cajun KnightHOST
51:14
Basically, Flatland has three shapes.
The Cajun KnightHOST
51:17
You've got triangles, circles, and squares, and it's all 2D.
The Cajun KnightHOST
51:20
So all you can really see is how a triangle starts at the point and it goes wide.
The Cajun KnightHOST
51:25
right and the circle kind of evenly goes around the squares flat and then you could see the edges kind of thing but that's all you could see one day a sphere enters the 2d realm and like fucks up this triangle who's like what am i even looking at because like a sphere passing through the 2d would look like a circle that got was small got wider and then got small again and He takes the triangle from the 2D realm and shows him the 3D realm.
The Cajun KnightHOST
51:50
It's really crazy.
The Cajun KnightHOST
51:52
The whole thing was meant to be like a political satire to the Victorian era England.
The Cajun KnightHOST
52:03
So if there's a third dimension, take me to see the fourth dimension.
The Cajun KnightHOST
52:05
And the sphere is like, there's no fourth dimension, you idiot.

Unknown podcast

Numerical Methods for Predicting Structural Stress

Aug 28 · 8 Mentions

Mason WilsonHOST
20:22
That's so clever.
speaker_1HOST
20:23
They modeled a 2D plane of solid material as an equivalent framework of elastic rods.
speaker_1HOST
20:29
It was a structural analog using a known structural framework to approximate an unknown continuum.
Mason WilsonHOST
20:35
It's an incredible hack.

13 MINS LATER

speaker_1HOST
34:00
The geometry and the physics of the problem dictate what kind of math you should use.
speaker_1HOST
34:04
The text breaks this library down into distinct dimensional categories.
Mason WilsonHOST
34:08
Like 1D, 2D, 3D.
speaker_1HOST
34:09
First, you have 1D line elements.
Christian SpicerHOST
20:20
There's bummer to it, but I don't wanna focus on the bummer side of it, if that makes sense.
Christian SpicerHOST
20:25
My story of the week is that 2XKO, which is Riot's, um, fighting game, tag fighting game, the, the, the 2D, you know, character fighting game that they had been working on for a while and was teased and, and slow re- It had a...
Christian SpicerHOST
20:39
Oh, what was it called? Project Alpha? Project...
Christian SpicerHOST
20:42
All right.
Luisa HoganHOST
0:01
If you've been solving the same problem for years and you haven't gotten anywhere, and you've thought that pushing harder was the way to tackle that, maybe the map you're using is the one that is holding you back, and harder is one of only two directions available.
Luisa HoganHOST
0:19
So my guest today says that the map that we're using is missing a dimension, and that we think in 2D when there's a whole new dimension to think about, and that most of us flatten our lives to fit in with... to fit in that 2D dimension.
Luisa HoganHOST
0:34
And he's already proven this with me because I have worked with this guest before, and still do, and I know that you are gonna absolutely love having him here today.
Luisa HoganHOST
0:45
So welcome back to the Second Voice podcast, where we uncover the conversations that leaders don't have out loud.
Luisa HoganHOST
2:19
Yeah.
Luisa HoganHOST
2:19
Let's get into it.
Luisa HoganHOST
2:20
So usually I start with like, you know, tell us more about you, but I actually wanna start with the whole 2D, 3D thing because-
Mark ZanGUEST
2:25
Sure

4 more episodes mention Two-dimensional space.

Create an account to see the whole feed, search across every transcript, and follow the entities you care about.

We value your privacy

We use cookies to understand how you use our platform and to improve your experience. Click “Accept All” to consent, or “Decline non-essential” to opt out of non-essential cookies. Read our Privacy Policy.