Refurio Anachro

I like to learn and popularize higher maths. I'm not a pro, but I can stomach corrections!

Codeberg: https://codeberg.org/refurio

Public Key

npub17lwktvf6fa9mlehaafftx3a78ge47js4pxd25qvm0wdl80dqtn7sde5ctv

NIP-05 Address

refurioanachro_at_mathstodon.xyz@momostr.pink

Profile Code

nprofile1qqs00ht9kyay7jalum77554ng7lr5v6lfg2snx42qxdhhxlnhks9elgpz3mhxue69uhhyetvv9ujuerpd46hxtnfduqs6amnwvaz7tmwdaejumr0dsy5mu46

Publishing to

relay.damus.io

nos.lol

Author Public Key

npub17lwktvf6fa9mlehaafftx3a78ge47js4pxd25qvm0wdl80dqtn7sde5ctv

Show more details

Last Notes

2026-02-26 18:43:06 UTC

- reply

For starters, the director Andrea Joshua Asnicar made a short series of short episodes exploring the scp scape.
https://www.youtube.com/watch?v=w-IiVeGAydE
@npub1nf4…nqe4

2026-02-20 16:01:47 UTC

- reply

I'd like to see the Fourier transform of Riemann zeta, too
@npub1nf4…nqe4

2026-02-20 15:57:05 UTC

- reply

I'm talking about the inverse of zeta for the primes 3 and 5 at the critical line. And I'm guessing the idea works for any set of primes.
Of course you knew what it is. I'm just stubbornly sticking to guessing because it's easier. Don't mind me %-D
I just wish I knew more about why the critical line is interesting. Your posts suggest one very hands-on idea, but I feel I might want to know about some of the countless other ideas.
And why am I not seeing any zeros? My fourier approximation has +1 at the end. Have I missed a -1 in ζ?
@npub1nf4…nqe4

2026-02-20 15:19:53 UTC

- reply

I've updated the shader to implement a factor of 1/√n instead of 1/(2log n) and the result is a much better fit. I'm guessing the fourier transform looks like this
1+Σ (-1)^k(q)•cos(x log(q))/sqrt(q)
summing over all q of the form 3^a•5^b.
I don't know what k(n) is, but it starts like 1,1,0,...
I'd also guess that this should generalize to the other rings of similar form.
@npub1nf4…nqe4

2026-02-19 04:46:12 UTC

- reply

Here's the shader for anyone to play with. It shows 1/ ζ, and that is what the cosines approximate.
https://www.shadertoy.com/view/w3VfRm
@npub1nf4…nqe4 @npub1ycr…856a

2026-02-19 03:05:44 UTC

- reply

For no good reason, I tried guessing the first few terms of the fourier transform myself.
(cos(x*log(3))/log(3)+cos(x*log(5))/log(5))/2+1
It's even better when we add
cos(x*log(15)-pi)/log(15)
I'm not sure where the extra phase pi comes from.
@npub1nf4…nqe4 @npub1ycr…856a

2026-02-18 15:13:39 UTC

- reply

I'm confused about this. The complex function has poles on the line Re(z)=0. There are no zeros on Re(z)=½.
The pattern you were looking for happens when two poles get close to each other. At any one pole (zeros, too) we see a dark and a light curve meet. When two poles get close it looks as if four lines are meeting.
In these cases one light image of the real axis points more or less straight to the right. Which means it gets closer to 1+0i faster. But since you were looking not at the poles there'd be an error depending on how straight that light curve is.
@npub1nf4…nqe4

2026-02-18 02:39:41 UTC

- reply

Looks like a somehow smoothed version.
@npub1nf4…nqe4

2026-02-16 23:39:30 UTC

- reply

Here's a complex plot up to 20i. You can see a unit circle at the bottom.
@npub1nf4…nqe4
https://media.mathstodon.xyz/media_attachments/files/116/082/937/144/338/811/original/8104550f0219b14c.png

2026-02-03 02:49:22 UTC

- reply

Here's Adam Neely about AI in music and the link to fascism.
https://www.youtube.com/watch?v=U8dcFhF0Dlk
As always, somewhat tangential to your thread, @npub1nf4…nqe4.

2026-01-28 07:04:43 UTC

- reply

I'd call that a good start!
@npub1nf4…nqe4

2026-01-28 06:20:36 UTC

- reply

Oh. Sure.
@npub1nf4…nqe4

2026-01-28 06:18:19 UTC

- reply

> I fixed the indentation
I ran it and it works fine now.
I'm glad you found a way to access programming for short computations! Claude sure takes away the burden to have all the vernacular ready. How easy do you find to read the result?
@npub1nf4…nqe4

2026-01-28 05:20:25 UTC

- reply

A bit more on-topic, I noticed that you mention Gene Ward Smith's work on classifying temperaments using wedge products (and Grassmannians)! That sounds curious, I'll have to read up about it!
@npub10sq…a55p

2026-01-28 05:01:16 UTC

- reply

You joke is quite funny. It made me look at your crackpot index once again! It's oddly specific at places :D
https://math.ucr.edu/home/baez/crackpot.html
How did you manage to come up with that python code? Looks fine to me. However, the indentation on azimuth looks broken in my browser.
@npub10sq…a55p

2025-12-23 17:39:07 UTC

- reply

Such a shame that photons aren't point particles. That would make for a fascinating lamp! Orb of wisdom or something.
@npub1nf4…nqe4 @npub1dxf…v5xz @npub12cu…d38p

2025-12-15 16:57:54 UTC

- reply

> That's because it's the volume, not the radius of a black hole that's proportional to its mass!
That's a typo. The radius of a black hole is proportional its mass. It's density is one over the square of either, whereas the density of a gas is constant.
@npub1nf4…nqe4 @npub1lez…3yeq @npub1dpn…wzrw

2025-12-13 10:01:22 UTC

- reply

Now here ^ is a nice start into your weekend: learn a bit about cospans!
The automatic subtitling is a bit obnoxious: "Marco spins" vs 'are co-spans'. Also: "coast-bands".
Not that you could do anything about it, @npub1nf4…nqe4. And thanks for your talk!

2025-12-08 14:54:18 UTC

- reply

So this is what you have been doing! Neat and crisp once more!
Huh, I made a screenshot of this bit:
> We can dramatize this. A Riemannian manifold M where each tangent space has a complex structure J: T_x M -> T_x M preserved by parallel transport is called a Kähler manifold. A Kähler manifold where each tangent space has a complex volume form ω preserved by parallel transport is called a Calabi-Yau manifold.
I wish I had that when I first felt I might want to learn what Calabi-Yau manifolds are about.
@npub1nf4…nqe4

2025-12-08 14:34:59 UTC

- reply

AI is not going to go away. I have mixed feelings about this. At work, I sometimes find myself spending hours cleaning up the mess others made with an assistant in just a few minutes. But there are good results, too! I think the balance is a bit off, though that depends on a lot of things. For example, it seems to be good enough for small things, if you don't care much where you end up.
Also: the hype!!
Do we need all this? Most certainly not. But what about the fun bits, can we afford those? Should we?
I believe, at this point, it's much more important to invest in formal methods than in AI search for theorems. Is that in fact the case? I don't have any figures. I'm not really afraid that AI is any threat to mathematics as we know it. But as I had intended to indicate earlier, it seems likely that a bunch of people will split off, maybe the gap to engineering will get bigger. Maybe we'll get to a different understanding of the achievement to find a proof.
Isn't it a good thing for mathematics to get better tools? We just have to figure out how to use those tools to do _real_ maths. I find it to be not much fun to talk to chat bots about maths, so that can't be it...
@npub1nf4…nqe4

2025-12-08 13:30:58 UTC

- reply

I agree with you, @npub1nf4…nqe4! I, too, feel helpless watching the systems we built to solve problems too big for any individual fail us at this.
The problem isn't criminals. It's that some groups of people aren't held accountable for their actions. All their claims of superiority, the gloating, the hate. These are clear signs of a damaged personality, who has lost access to very basic modes of cooperation. No matter the amount of philanthropy, they won't help us solve the real issue. They are not part of the solution!
What we are doing right now amounts to an elementary economic misjudgement. Successful cultures would plan ahead. Some failed. Maybe we will turn out to be one of them. But I have not lost all hope. Most of my life there was this dread and a feeling of being stuck. But things are changing! I want it to happen faster and it's no fun to see the momentum swinging back. But I think that is to be expected and will only intensify the next wave forward. We need to keep pushing! I certainly will.

2025-12-07 22:34:18 UTC

- reply

Software verification is a big one. I believe that we will eventually need verified software everywhere to help solve fundamental security issues we have today. And indeed, there's so much tedious work involved with current tools that it's simply infeasible beyond very small code bases.
I love Terry Tao's approach! It strikes right at the heart and effectively makes weaseling much harder. I enjoy reading his finds a lot. It's good to see he gets support by a community.
Research, teaching and engineering are intertwined. At least to Archimedes' times they were, he was a carpenter and architect. Of late, the engineers seem to have been incing away a little. Maybe to catch some new fish. If you had an oracle, you would use it, wouldn't you?
@npub1nf4…nqe4 @npub1fp0…s4sx

2025-11-23 15:04:59 UTC

- reply

I wondered about the ordering, too. That group of parallel slanted lines irritated me. But then I realized that the numerical sorting is point symmetrical, which made me feel slightly better. It may be easier to change the ordering in my .awk program, so if you have any suggestions...
Also, you do know how to change the shape (or color) of the nodes, do you? At the moment it has sort of a brutalistic appeal, but that can easily be changed, and, right now, I happen to know how.
@npub1nf4…nqe4 @npub1p5z…4d3n @npub1t2k…2wfh

2025-11-22 21:46:40 UTC

- reply

I just checked, it runs with pdflatex again. Just remove the `\usetikzlibrary{graphdrawing}` bit (leave `graphs` in there).
Some of the layout algorithms are written in lua, but we have coordinates now.
@npub1nf4…nqe4

2025-11-22 18:48:51 UTC

- reply

My explanation makes no sense. I can still imagine the option (or processing step?) "linear optimization node positioning", which tries to draw edges as vertical as possible, to be responsible, but that may also make no sense %-)
I couldn't get the algos to do the placement correctly, the relative manual placement methods felt confusing, and since I'm generating TikZ anyways (in case someone needs a 7-cube to fill pages), it's much easier to simply compute coordinates.
Hope it's a bit more useful now.
https://paste.fyi/maYyxcge
@npub1nf4…nqe4

2025-11-21 21:41:39 UTC

- reply

Yes, the placement is probably due to the algo avoiding one edge crossing in each row. That can probably be turned off. I'll have a look later if I can find out how.
@npub1nf4…nqe4

2025-11-21 20:38:31 UTC

- reply

Looking at how much fun I just had to get TikZ do what I want, I hope it's okay if I share a few more lines, so here's what TikZ's layered graph layout algorithm makes of the graph:
https://paste.fyi/GJMDfXqe
main.tex:
\documentclass{article}
\usepackage{tikz}
\begin{document}
\input{5-cube.tikz}
\end{document}
$ lualatex main.tex
I still feel like a total TikZ noob. I suppose to change labels, you'd not draw labels but then set [label=something]? Everybody: feel free to use this in any way.
@npub1nf4…nqe4 @npub13gv…v40m
https://media.mathstodon.xyz/media_attachments/files/115/589/545/785/717/856/original/a720f24f637c4836.png

2025-11-20 18:33:35 UTC

- reply

I have no TikZ here, and no idea how to use it, but would this help, @npub1nf4…nqe4? I might get back to it later and see what options for layouting there are.
v00000 -> v10000;
v00000 -> v01000;
v00000 -> v00100;
v00000 -> v00010;
v00000 -> v00001;
v10000 -> v11000;
v10000 -> v10100;
v10000 -> v10010;
v10000 -> v10001;
v01000 -> v11000;
v01000 -> v01100;
v01000 -> v01010;
v01000 -> v01001;
v11000 -> v11100;
v11000 -> v11010;
v11000 -> v11001;
v00100 -> v10100;
v00100 -> v01100;
v00100 -> v00110;
v00100 -> v00101;
v10100 -> v11100;
v10100 -> v10110;
v10100 -> v10101;
v01100 -> v11100;
v01100 -> v01110;
v01100 -> v01101;
v11100 -> v11110;
v11100 -> v11101;
v00010 -> v10010;
v00010 -> v01010;
v00010 -> v00110;
v00010 -> v00011;
v10010 -> v11010;
v10010 -> v10110;
v10010 -> v10011;
v01010 -> v11010;
v01010 -> v01110;
v01010 -> v01011;
v11010 -> v11110;
v11010 -> v11011;
v00110 -> v10110;
v00110 -> v01110;
v00110 -> v00111;
v10110 -> v11110;
v10110 -> v10111;
v01110 -> v11110;
v01110 -> v01111;
v11110 -> v11111;
v00001 -> v10001;
v00001 -> v01001;
v00001 -> v00101;
v00001 -> v00011;
v10001 -> v11001;
v10001 -> v10101;
v10001 -> v10011;
v01001 -> v11001;
v01001 -> v01101;
v01001 -> v01011;
v11001 -> v11101;
v11001 -> v11011;
v00101 -> v10101;
v00101 -> v01101;
v00101 -> v00111;
v10101 -> v11101;
v10101 -> v10111;
v01101 -> v11101;
v01101 -> v01111;
v11101 -> v11111;
v00011 -> v10011;
v00011 -> v01011;
v00011 -> v00111;
v10011 -> v11011;
v10011 -> v10111;
v01011 -> v11011;
v01011 -> v01111;
v11011 -> v11111;
v00111 -> v10111;
v00111 -> v01111;
v10111 -> v11111;
v01111 -> v11111;

2025-11-19 12:44:26 UTC

- reply

I suppose that means you won't run out of knotty jokes anytime soon.
I'm still amazed that you guys made the move. It must have been difficult. I didn't have Edinburgh on my map, from what you have shared it seems to be a very nice and also somewhat quiet place. And much more maths history than I would have suspected from, uh, a blank spot on my map. University founded in 1582, that's not shabby at all!
@npub1nf4…nqe4

2025-11-19 12:31:05 UTC

- reply

> Edges […] are strings with one of the bits replaced by a \*. Squares ...
Clearly. I wasn't aware. Cool, thanks, @npub1p5z…4d3n!
An edge would then describe a pair of processes taking either value of a property to its opposite.
I suppose I'm stretching the model too much with faces. My motivation was to describe other topological obstacles there might be.
If we don't use its symmetries, introducing hypercubes seems -needlessly complicated- a good opportunity to talk about higher-dimensional stuff.
I'm glad you noticed I have no idea what I'm talking about :D
@npub1nf4…nqe4

2025-11-18 18:16:23 UTC

- reply

Which of the cube's edges are present? Are faces missing?
@npub1nf4…nqe4 @npub1y7v…jtwa

2025-11-06 13:34:29 UTC

- reply

Proof idea: Take a square touching a circle at four points. There's one degree of freedom, the square's orientation. Here, the conditions, tangency and that the square's edges midpoints are hit, hold trivially. We can now squeeze the circle to get an ellipsoid. This generalizes easily.
Not quite sure if I'm missing anything. Should I also prove that we can get the same set of parallelograms that way?
@npub12cu…d38p

2025-10-28 18:20:34 UTC

- reply

> You're also mixing up "Cauchy surface" and "Cauchy horizon", which are completely different.
Ah, there it is. The place where my confusion starts. It's been with me for a while! Thank you :D
Why is it called globally 'hyperbolic'?
> a Lorentzian manifold containing a Cauchy surface is said to be globally hyperbolic.
So a single Cauchy surface implies it can be foliated by Cauchy surfaces!
> Globally hyperbolicity excludes closed timelike curves, but also other bizarre behavior.
Hm... 'other bizarre behaviour'. Since g(v,v)>0 this could be tachyons? Are singletons (g(v,v)=0) a thing?
Or separated regions to which S somehow cannot be extended. Or... Cauchy horizons? No, that is just guessing, I haven't quite finished reading about Cauchy horizons.
@npub1nf4…nqe4

2025-10-28 16:12:15 UTC

- reply

>> The existence of a naked singularity implies a failure of global hyperbolicity.
Hyperbolic in the sense of a Minkowski manifold, which basically means we cannot rotate (boost) an inertial frame to point outside its future light cone. Global means that I can't get outside my future lightcone even by topological means allowing me to travel back in time.
https://en.wikipedia.org/wiki/Globally_hyperbolic_manifold
They have definitions with funny mames:
> M is non-totally vicious if there is at least one point such that no closed timelike curve passes through it.
Or:
> M is non-total imprisoning if no inextendible causal curve is contained in a compact set. This property implies causality.
And an old acquaintance, the Cauchy horizon:
> A Cauchy surface for M is a closed achronal set whose domain of dependence is M.
That is a place where the future development is not sufficiently determined by the laws of physics!
Your link makes it very easy to spend my afternoon reading your paper, instead of following my day-job. Very tempting indeed!
@npub1nf4…nqe4 @npub100e…c8sg

2025-10-28 14:39:29 UTC

- reply

> Hawking proposed the cosmic censorship conjecture that no naked singularities form in GR given realistic starting conditions. There is evidence from numerical simulations that this is false.
What?! I didn't notice this happened! Has the community already turned away from the cosmic censorship conjecture? Or are there good reasons to mistrust these numerical simulations? Is this a recent result?
On my todo list I have that I want to make a shadertoy with relativistic raymarching. I suppose I now want to illustrate a fall into a white hole! Colliding things is a bit harder, not just because the computation is more inolved, but also because the shader language was made to render textures, so it's not up to iterating over a large chunk of data.
Anyways, thank you very much for that information, so exciting!
@npub100e…c8sg @npub1nf4…nqe4

2025-10-27 23:47:05 UTC

- reply

Oh wow, this is really fun to think about. Two white holes can't merge, because the time reverse would be a black hole splitting up, and that's not allowed!
Your sketch for colliding black and white holes is interesting. It seems almost symmetrical. They both stick together until the white hole unforms and then we get our black hole back.
@npub100e…c8sg @npub1nf4…nqe4

2025-10-27 21:57:45 UTC

- reply

Whatever the outcome, it should be time-reversible, right?
@npub100e…c8sg @npub1nf4…nqe4

2025-10-27 21:02:09 UTC

- reply

Watch this cool interview for a nice discussion about white holes! I had reshared this earlier, but now that I've seen it: it's really good!
https://mathstodon.xyz/@johncarlosbaez/115445374242954199
@npub1nf4…nqe4

2025-10-27 20:58:49 UTC

- reply

> might have [met]
Apparently not. Nevermind, @npub100e…c8sg :D While we're at collecting questions, what happens when a white hole and a black hole collide? Looking at the Penrose diagram it seems like they can't merge?
@npub1nf4…nqe4

2025-10-27 12:30:00 UTC

- reply

Do you happen to be a (former?) collector of VAX machines, @npub100e…c8sg? In that case we might have had a chat on-line after John gave a talk a few years ago. If so, cheers!

2025-10-25 18:07:06 UTC

- reply

I now did. We learn that since category theory turned out to be such a failure for physics we should just wait and see the same happen with positive geometry. To be fair, she argues pro mathematics, but does so while holding a book into the camera written by a guy arguing the other way.
It's... Am I trolling. Maybe. I actually like her final remark: if you want to improve your critical thinking, go to Brilliant.
It's clearly more fun to get on and learn something. Indeed, I prefer listening to people who do that in a positive way.
Here's Alexander Postnikov's paper about positive Grassmannians:
https://arxiv.org/abs/math/0609764
It's quite good!
@npub1nf4…nqe4 @npub1my0…d6ft @npub1rsq…33yg

2025-10-23 17:00:20 UTC

- reply

Yeah, I noticed you guys having fun juggling Jordan algebras, you even got me to fresh up my memory about them! I have no connection to them, so I guess I have some catching-up to do. Sometimes, my understanding gets better when I don't look at something for a while, but I need some sort of intuition for that to work, and I don't have much for Jordan algebras. Is there a geometric interpretation for them?
@npub1nf4…nqe4 @npub1zz2…9mhx

2025-10-22 21:20:38 UTC

- reply

> You can watch my slow technical progress.
@npub1nf4…nqe4
A friend recently got agitated, 'Why, yes, I'm a kind of neanderthal, you can watch me learn!'. I can identify with that! Of course, it's from a corporate setting, where people are more often reluctant to admit they have no idea about the matter at hand.

2025-10-22 21:15:33 UTC

- reply

Getting to read mathstodon from my phone again should make a difference. I hesitated to move because I've had enough of political social media. That other account of mine will have to sleep a little longer, I suppose. Can't live without the maths, though.
And haven't, really. There's Fractal Chats on Discord, @npub1kr6…lhf0's HyperRogue Lounge has a nice hyperbolic geometry channel. Noticed I forgot how to compute cohomology...
I'm also making interactive maths in 60 frames per seconds on shadertoy.com. Even released a 4k intro at the Evoke demoscene party. That got me in touch with some video artists who keep imploring me to put up visuals at parties. Makes me feel old and young at the same time.
@npub1nf4…nqe4

2025-10-22 20:24:54 UTC

- reply

I see you are getting better at this already! Cheers!
@npub1nf4…nqe4

2025-09-10 21:50:13 UTC

- reply

Ah, the good old exponential \king^\queen.
@npub196u…kntt @npub1qhm…qhxt @npub1nf4…nqe4

2025-05-09 15:12:44 UTC

- reply

123123 * 5^2 = 3078075
In particular, 078 != 075
@npub12cu…d38p

2025-03-23 04:15:28 UTC

- reply

Abdelaziz Nait Merzouk made gorgeous renderings! They are really something to aspire to. I would really like to make similar ones in shadertoy. First get a Newton root finder implemented and then see what I can do with that. Another inspiration is to try and get something like [@KangarooPhysics](https://cornerof.world/users/KangarooPhysics )'s Seifert-like surfaces for knots.
Another project idea is to write a relativistic raymarcher, to render black holes and some other, more weird 4d stuff.
I'm kind of obliged to something along these lines, because I promised some guys to show off with them at the next Evoke demo party in Cologne. That means I have about half a year to get it done, which should be plenty. Should...
@npub1nf4…nqe4

2025-03-23 04:07:39 UTC

- reply

I finally managed to watch your talk, @npub1nf4…nqe4! Those were good times, and I really enjoyed going through your visual insights again, looking at all those nice pictures and remembering the things I learned. Or more like rediscovering them.
You said that you'd like to have another, better look at that spheroid. Well, perfect spheres are rather boring. I've attached an animation I captured from my shadertoy implementation. The point where it gets quite symmetric is close to what a sphere looks like. The browser version allows you to change its parameters to perfectly spherical by clicking onto it. Left/right changes the spheroid's shape, up/down the position of the light source.
This version doesn't behave well in more extreme configurations, black means numerical instability, red means it's too chaotic for anti-aliasing to work. The color changes from orange to blue and back every other iteration. If I'm not mistaken the orange parts are odd, and the blue parts have undergone an even number of reflections. Deep blue means the light source wasn't hit after 50 iterations.
All parameters can be changed either at the top of the file, or down in the MainImage1 function. If you want to animate the angle you need to remove "//" from line 151, and add "//" to line 158, so the animated parameter isn't overwritten by the mouse cursor.
That external ray pattern shows the slope of the spheroid in the picture plane, I added this for debugging purposes. It all looks right, but I'm not entirely sure I made no mistakes. I'll certainly have to spend some more time to improve on the black parts and to investigate the strange pictures for more extreme parameter settings.
https://www.shadertoy.com/view/43fcRs
https://media.mathstodon.xyz/media_attachments/files/114/209/728/803/583/882/original/2d0e4b0d13a11c20.mp4

2025-03-15 19:02:29 UTC

- reply

Reminds me of when my fridge's thermostat froze, after my cleaner had accidentally turned the dial to 11. Pulling the plug and waiting a few hours worked at the end. Which took me almost a week to grok. I now know much more about handling frozen cheese than I ever cared to.
That genius of a person had also rubbed all of the inside with loo cleaning agent. The remedy here was much more difficult and involved: cleaning with vinegar helped somewhat. At the end, putting in lots of rumpled newspaper, and waiting for a week or so did the trick.
@npub12cu…d38p

2025-03-15 14:05:03 UTC

- reply

That looks like a lot of fun. I aimed for it, but couldn't make it yesterday. I'm looking forward to see the recording!
@npub1nf4…nqe4

2025-03-06 03:08:07 UTC

- reply

I suppose you can be good at something and still have a black spot in your eye. I haven't read Sabine's GR papers yet (thanks for linking), but I have no trouble believing that she knows a lot about GR. I should have been a bit more careful in how I put things earlier. After all, I greatly enjoyed reading her work on superdeterminism, she's clearly a capable researcher!
The point I won't change my mind about is the twin paradox: it's not caused by acceleration! But I can see that, if we get very philosophical, we can pin the cause of an effect on something that's correlated with it, if that's always the case, because we've limited the allowed cases such that that works out... I remember Greg Egan saying that this is not "terribly misleading", and that made me think a lot deeper about this. So I have to agree, the paradox can be put in a way that it is technically acceptable to say it's caused by acceleration. But it's morally wrong! It makes understanding space-time metric harder!
I couldn't let it pass, but I could have done better, so... thanks, John!
@npub1nf4…nqe4 [@j_bertolotti](https://mathstodon.xyz/@j_bertolotti ) @npub1p7u…kglz @npub1pkc…qy0t @npub1x78…xn0m @npub1glg…tqzy

2025-03-06 01:59:50 UTC

- reply

To be fair, the stuff I feel comfy with may not have much overlap with her area of expertise. Which could explain why things look more bleak where she talks about stuff I spent a lot thinking about. For the record, though, I'd have to formally contest that she's well versed in relativity.
Anways, we don't have to compare bad with worse. Let's instead emphasize those who do better! Maybe like so:
https://mathstodon.xyz/@RefurioAnachro/114095112028682279
[@j_bertolotti](https://mathstodon.xyz/@j_bertolotti ) @npub1p7u…kglz @npub1pkc…qy0t @npub1x78…xn0m @npub1nf4…nqe4 @npub1glg…tqzy

2025-03-02 23:36:53 UTC

- reply

I just tried to search for Jim's lecture, but that's not easy, because he's not the famous billionaire mathematician with the same name! He's that other one I like so much #-/
Well, I think it's not in his series about the history of trigonometry, but I think it was in his nice feature about log tables:
https://www.youtube.com/watch?v=9D3Rp3oiMVQ
Thanks for your thoughts, appreciated as always, @npub1nf4…nqe4! Hey, you also have a Youtube channel that found its way into my bookmarks! Even though your regular cast with James Dolan seems to be more for fans, there are some gems in between! I'll see into mentioning them in my next batch!

2025-03-02 22:13:33 UTC

- reply

Jim Simons likes to talk about maths that might fit well into high school curricula. But he has an eye for details that most courses don't find patience for. Did you know that people, before the advent of calculators, before they used log tables to simplify multiplications, they used cosine tables to do the same? Jim goes deep, demonstrates how things work, and is not afraid to clad teaching material in new ways.
https://www.youtube.com/channel/UCnYszOhEIIdIMYyNx2yjwfg
You will need to concentrate a bit, but then it's a really entertaining and fun channel!

2025-03-02 22:07:27 UTC

As promised, here some commented links to Youtube channels I like to watch. Most of them are about maths or astro, but not all of them. I am intentionally not starting with the most popular ones. Oh, and I'd love to hear about your favorite channels, too! If you write a few lines about your link, I'll reshare your post.
Chris Staecker is a collector of computing devices. He demonstrates them with a pinch of nerdy humour, and boy has he some awesome gimmicks to show. More often than not you will need some historical context and Chris delivers.
https://www.youtube.com/channel/UCpFN5_YDQpHMF2XGpV9JKrA
He is also doing research on discrete topology and giving classes, and those videos might also end up in his stream. Give those a try, too, they are just as well done.

2025-03-02 21:58:00 UTC

- reply

Let me put it this way: there are certainly episodes that are less awful than others. However, the more I know of a subject the worse they tend to look to me. Like the one where she got quite exhilarated over a paper showing that nature can behave non-computable. Or did it? Well, we prove this sort of thing for, say, a formal system, by showing that we can implement a computer in it, and then point at the halting problem. To be fair, the paper established that some quantum effect is Turing complete, not that "nature is non-computable". Because the latter wouldn't be news to anyone at all. Just look at any computer next to you...
I guess that she doesn't do much research on most of the stuff she talks about, and her team really isn't up to it. Their selection of topics looks a bit as if it was taken from a conspiracy channel on Telegram, or some neoliberal tech outlet. Just look at the sheer amount of AI, quantum computing, or cryptocurrency hype they serve. I'm not against trolling per se, but getting superficial info drilled in as if it was the ultimate truth isn't something I enjoy much. Especially not if it's politically shallow.
Anyways, it's good to see some of the cool people starting to get disenchanted (thanks for that word, Bartosz!). I must be extra sensitive, the first time her channel put me off was a while ago. I wish I could buy something for that trend sensing. Or, forget it, I don't care about glamour or sex appeal, pay me by explaining theorems instead. Or chocolate works, too.
@npub1p7u…kglz @npub1pkc…qy0t [@j_bertolotti](https://mathstodon.xyz/@j_bertolotti ) @npub1nf4…nqe4 @npub1glg…tqzy

2025-03-02 20:56:49 UTC

- reply

There are good channels, too. I guess I'll have to make a list.
@npub1pkc…qy0t @npub1p7u…kglz [@j_bertolotti](https://mathstodon.xyz/@j_bertolotti ) @npub1nf4…nqe4 @npub1glg…tqzy

2024-12-26 18:30:40 UTC

- reply

> I can argue that the fact that it's not a point source makes it less bright when you get very close than the inverse square law predicts, not more. I can also argue that it has no effect. So I'm not sure.
Good point! This is more fun to think about than I had noticed at first!
@npub1nf4…nqe4 @npub10hg…t88z

2024-12-26 16:11:49 UTC

- reply

Shouldn't it be 25²=625 times brighter? Maybe that's not quite right, because it's not a point source it should also look bigger and that makes even more light reach Parker.
@npub1nf4…nqe4 @npub10hg…t88z

2024-11-29 12:32:15 UTC

- reply

It gives a comprehensive view of a supernova and I got caught by its nice diagrams. It's still lying on my desktop, but I'm not sure when or if I'll make it back to reading it. Glad you enjoyed taking a second look!
@npub1nf4…nqe4 @npub1gex…rjyz

2024-11-20 01:30:24 UTC

- reply

This paper has much more to offer than estimates on supernova kicks! They implemented reaction networks to model nucleosynthesis! This is fun! They even show pictures of this most extreme stellar process, so folks, go look at it! The scales are indeed mind-boggling.
> Natal
Aww ❤️
So, The kick flips it over, and then the binary partner winds it up? Or how does it happen that the magnetic and rotational poles diverge?
Anyways, neat find! Thanks! You're right, I should not shy away from searching for papers. My clock is broken, the time it shows isn't plausible.
@npub1nf4…nqe4 @npub1gex…rjyz

2024-11-19 21:45:02 UTC

- reply

Supernovae are violent processes where most of the star's mass gets ejected at more than escape velocity. This can happen in an asymmetric way, such that the remnant itself can get whipped away at high speed. I find it highly plausible that the star might pick up angular momentum. I'd like to support this hypothesis with papers, but I'm not used to finding them, and I'm feeling lazy.
Another idea is that, often, the magnetic and rotational poles of a pulsar don't match. The lighthouse beaming follows the magnetic poles, sometimes there are even more than two! Anyways, that's how we get pulsars, if these were always aligned, there would be no blinking!
@npub1nf4…nqe4 @npub1gex…rjyz

2024-10-20 22:29:31 UTC

- reply

> Philip Anderson [...] 'effective' theories.
Thank you! I felt that was an excellent piece of trolling, at least the way you put it. I mean, in a constructive way. It should be fun to look up what Anderson said exactly. We should be looking at QFT beyond its applications in fundamental physics! And I have a gut feeling that that may pay off in unexpected ways, perhaps even surprising connections between the realms the QFT's apply to.
At least to encourage people study QFT's in a more general setting. When I finally got around to learn about classical spinors (Pauli, Weyl, Dirac), I was quite shocked how much of a cludge they are, compared to the more general Clifford setting.
> There's a lot of real physics in quantum computing, regardless of whether it pays off soon or ever.
Certainly. I remember following Scott Aaronson's development of his position, wether quantum computing is practical, anxiously. But even so, there's lots of hype, and I just claimed that that's what the Nobel committee is up to...
BTW, I really like watching Nima's lectures. So thanks for linking one, I hadn't seen one in quite a while.
@npub1nf4…nqe4

2024-10-20 19:32:06 UTC

- reply

I wouldn't know what Sabine said, she's a much more proficient troll than me, and I kept getting tangled up with what she said, so I stopped watching her videos. But I heard she's been experimenting again, do you think I should give her show another chance?
In any case, it was you quoting someone saying something along the lines that material scientists might not agree that fundamental physics is so fundamental after all, right? That is a very deep idea and I keep thinking about it! It would be great if there was a periodic table of quantum fields, maybe cross referenced with metamaterials giving rise to them. In any case, material sciences getting more attention? I'm all for it!
Quantum computing on the other hand is attracting lots of venture capital, I do wonder if that will pay off anytime soon. But it might be a good candidate for continuing the pattern. I mean, the Nobel prize one, not the grift one, as in Bitcoin, NFT, AI. Sounds like an interesting bet anyways: the next physics Nobel, will it be about quuantum computing?
@npub1nf4…nqe4

2024-10-20 15:28:04 UTC

- reply

In that light, here's my suggestion: After honoring Roger Penrose last year, who, don't get me wrong, certainly deserved any prestigious award, he's still much more of a mathematician, so something like the Fields Medal might have been much more fitting, even back when he was young enough to fit their profile. But nevermind, Ed Witten got that one, one of the most intimidating theoretical pysicists you might imagine, and not a mathematician.
So this is the second year in a row where the Nobel for physics didn't go to a physicist. Maybe the reason is simply that physicists have stopped making progress. There is just nothing to award for, so of course the commitee reached out to adjacent professions. It's either that or not award the prize at all!
2/2
#NobelPrize #NobelPhysics #Nobel2024

2024-10-20 15:27:07 UTC

The controversy around this year's physics Nobel seems to be still on-going. See, for example, the joint release of a reaction video on Computerphile and Sixty Symbols:
https://www.youtube.com/watch?v=itywoDU3sL8q
(by @npub1mah…0g6q, and @npub1pxq…kcy3)
Wherein both sides express their surprise between the lines. But you have to take into account that the Nobel commitee always and above all has been celebrating popularizers, they never intended to emphasize the first person to come up with a most influential idea, nor the one who did all the work..
That's likely the reason why so few women got awarded this prestigious prize, as it is even more of a privilege to be in a position to popularize science, hard working female scientists rarely attain. See here for a take by Angela Collier towards that end:
https://www.youtube.com/watch?v=zS7sJJB7BUI
(by @npub1xzw…l6gs)
Or my friend John Baez' take, he also suggested to maybe tone down the scandalizing.
https://mathstodon.xyz/@johncarlosbaez/113272834785880929
(by @npub1nf4…nqe4)
Which is certainly going to work. On social media. Telling people not to get agitated. I'm sure it came as a total surprise when people expressed their feelings even more vigorously in response.
1/2

2024-10-05 17:42:01 UTC

- reply

> "TÜV Rheinland’s Alexander Schneider directed me to a digital certificate, which contained an almost complete absence of detail about how it was actually tested."
The relevant body is BSI, "Bundesamt für Sicherheit in der Informationstechnik", but it's not better with them. You can get a BSI certificate simply by claiming you did everything right, whatever that's supposed to mean :D
@npub12cu…d38p

2024-10-03 10:07:27 UTC

- reply

Fascinating, [@johncarlosbaez](https://mathstodon.xyz/@johncarlosbaez )!

2024-09-14 01:01:35 UTC

- reply

Aww. These shadertoy.com links should open to an animation. Firefox doesn't work for me, but Chrome does. Both should work under Windows. I remember I had to enable something on my PC, but it worked out of the box on my Android mobile.
@npub1nf4…nqe4

2024-09-14 00:01:28 UTC

- reply

I hope the animation does work for you, @npub1nf4…nqe4?

2024-09-13 23:43:45 UTC

- reply

Those were the days, @npub1nf4…nqe4! I really enjoyed working on this one! And now it's there for everyone to play with, and I don't have to make variations for other people anymore... I'm free :D

2024-09-13 23:27:48 UTC

Another shadertoy: "Sphere in Mirrored Spheroid"
https://www.shadertoy.com/view/MXfyDM
This one featured 2015 on @npub1nf4…nqe4' AMS blog "Visual Insight"
https://blogs.ams.org/visualinsight/2015/04/15/sphere-in-mirrored-spheroid/
Back then it took tens of minutes to get a nice picture. I spent hours staring at this thing, so it was certainly worth a remake. This new raytracer's quality is a bit lower than in John's blogpost, mostly because I haven't implemented anti-aliasing (supersampling). But boy, it is *fast*!
Look at the bottom of the source for some more ideas to animate.
https://media.mathstodon.xyz/media_attachments/files/113/132/804/338/900/810/original/32d7135beea4f36e.png