j_bertolotti

Associated Professor of Physics at the University of Exeter.
Maker of typos.
Science, scientific visualizations (grouped under the hashtag #PhysicsFactlet), some miniature painting.
He/lui/on. All opinions are my own fault.

Public Key

npub1a4ywuu668ac6558wk53ng6nsykluuag9zpwm64nhgmjpeucc5n6skrazsc

NIP-05 Address

j_bertolotti@mathstodon-xyz.mostr.pub

Profile Code

nprofile1qqsw6j8wwddr7ud22rht2ge5dfczt07ww5z3qhda2em5dequ7vv2fagpz4mhxue69uhhyetvv9ujuerfw36x7tnsw43qz9thwden5te0wfjkccte9ekk7um5wgh8qatzte5grf

Publishing to

relay.ditto.pub

relay.mostr.pub

Author Public Key

npub1a4ywuu668ac6558wk53ng6nsykluuag9zpwm64nhgmjpeucc5n6skrazsc

Show more details

Last Notes

2026-10-06 10:27:04 UTC

The #Nobel Prize in #Physics has been awarded to Francis Halzen “for decisive contributions to the IceCube Neutrino Observatory and the discovery of high-energy neutrinos of astrophysical origin”.
Obviously congratulations to him, but it always feel weird to award a single individual for the results of a collaboration between thousand of scientists.
https://www.nobelprize.org/prizes/physics/2026/halzen/facts/

2026-09-03 09:34:11 UTC

New paper on #Arxiv and my first foray into the field of #OpticalComputing (and #MachineLearning , the good one, not LLMs&co).
We are very used to digital universal (i.e. Turing complete) computers, but it isn't written anywhere that all computation must be digital. Actually, analog computers have a long and glorious history. The advent of transistors meant that digital computers won the war and became ubiquitous, but if instead of a universal machine you want something that does exactly one thing, but very fast and very efficiently, analog systems are often the way to go.
In particular it is tempting to do computation with light. Light is extremely good at transmitting information, we have excellent technology for both generating and detecting light, and we are VERY good at manipulating it. Furthermore light is FAST, and if you do things correctly losses are low.
And if there is something that would benefit from being made massively parallel, fast, and with limited energy consumption it is machine learning. Now, not all forms of machine learning suit light well. People try to make deep neural networks using light, but I am not convinced that this approach has legs. A less popular approach to machine learning that is a perfect fit for light and optics is "reservoir computing", where you have a large recurring network where all the weights (i.e. how strong the connections between the nodes are) are fixed, and you only train a single linear layer (effectively you just do a linear regression) at the output of the network.
1/3
https://arxiv.org/abs/2609.02733

2026-08-12 18:22:54 UTC

94% #SolarEclipse as seen from #Exeter (UK)
https://media.mathstodon.xyz/media_attachments/files/117/083/918/614/505/561/original/1b780096c4f715d9.jpg

2026-08-02 17:25:40 UTC

Every Waaagh must begin somewhere. Mine is beginning here.
#MiniaturePainting #WH40k #orcs
https://media.mathstodon.xyz/media_attachments/files/117/027/069/494/673/193/original/f7d07323339ddbdb.jpg
https://media.mathstodon.xyz/media_attachments/files/117/027/069/785/801/257/original/57230325f5854a24.jpg

2026-07-14 13:33:37 UTC

- reply

@nprofile…yx65 They say they will use AI "responsibly", so there is nothing to worry about!
/s
https://media.mathstodon.xyz/media_attachments/files/116/918/579/540/948/113/original/3fcd126c3a1c8d56.png

2026-05-14 10:48:28 UTC

#PhysicsFactlet
The "stretch and fold" dynamics of the phase-space representation of a chaotic system (in this case a periodically driven simple pendulum) is always fun to see.
(Made for a lecture, so I am quickly putting it here. Maybe will write a thread about it in the near future.)
#Chaos
https://media.mathstodon.xyz/media_attachments/files/116/572/521/248/438/249/original/5f4d9a07e7891fa7.mp4

2026-04-30 14:51:50 UTC

"The Mastodon time line is empty and boring"
Listen, I posted the picture of a couple of poorly panted Warhammer miniatures, got a few likes, looked at who liked it and followed those who seemed to boost a lot, and now I am awash of photos of gorgeously painted miniatures!
Your timeline is as alive (or dead) as you make it!

2026-04-09 10:55:24 UTC

- reply

@nprofile…np45 Just think about all the implied knowledge in any cookbook. Recipes can be compressed to a page or two only because they assume you already know how to do a lot of the stuff they are talking about. If you have no idea how to cook, no amount of cookbooks will ever help you.
@nprofile…jnc7 @nprofile…ugpy

2026-04-04 15:51:55 UTC

One more finished model, a million new things I tried for the first time. The most notable were blending contrast paints (the red on the wings and black on the blade on top of the head), and the crackling on the base.
Biggest mistake was not to undershade enough before applying the contrast paints.
#PaintingMiniatures #Miniatures #Warhammer #WH40k
https://media.mathstodon.xyz/media_attachments/files/116/347/215/107/425/718/original/5fd294674c8c1b0c.jpg
https://media.mathstodon.xyz/media_attachments/files/116/347/215/104/061/821/original/4dc39f9eb91b2f5b.jpg

2026-03-03 11:56:47 UTC

I see a lot of people posting about the Greens overcoming Labour in the polls, but everything I see in those polls is Reform being the first party.
And that is f***ing scary.

2026-02-13 10:12:20 UTC

Random reminder that, despite what you read in "Lord of the Flies", when a real group of teenagers happened to get stranded on a inhabited island, they didn't start killing each other. On the contrary they worked together, built a commune, and survived in good health for 15 months (when they were saved by a passing ship).
Most people are good people. Even teenagers 😉
https://en.wikipedia.org/wiki/Tongan_castaways

2026-02-02 16:15:24 UTC

- reply

@nprofile…4hp5 @nprofile…3vpz Counterpoint: noting in the game is photorealistic, so you might want to make the mountains less "realistic" to better fit with the general vibe.

2026-01-20 09:26:04 UTC

My long trip to Glasgow begins with me getting soaking wet on my way to train station. ☔
How can you not love the British weather?

2026-01-16 14:09:54 UTC

- reply

@nprofile…llgf Spoiler: we failed (current government is openly fascist).
After Berlusconi we had a couple of "technical" governments, i.e. prime ministers not really belonging to any party, but accepted by both sides who tried to bring the economy back from the brink of collapse. They sort of succeeded, but during this time the right had the time to complete their internal fights and establish a new pecking order, while the "left" largely disappeared. As a result, once the economy recovered a little bit, the right went back in power, with the racist and corrupt stuff that were made normal under Berlusconi now the norm.
If you want to avoid that, you need jump-start the Democratic party, kick out all the old guard who will always choose apathy, and rally around a new guard with progressive ideas.

2026-01-16 09:45:38 UTC

Dear US people, I am afraid I have bad news. Even if Trump were to disappear today, your problems would be far from over. Your society is in pieces, and it will not magically go back to "normal". The bar for what it is acceptable has now been moved, and the far-right vultures will take advantage of it.
How do I know? I lived under Berlusconi. And Berlusconi was a (smaller scale) precursor of Trump. You are following exactly the same path we did, just much bigger and much worse (you always do things bigger in the US), and thus it is not difficult to see what is coming for you. We know. We have already gone through it. We warned you and you laughed at us.
Please, stop laughing and learn from our mistakes. You are still in time.

2026-01-10 10:06:00 UTC

- reply

@nprofile…97ud I feel you have it upside down: AI might be get good enough to automate the process of getting "correct" Maths. What we will need Mathematicians for is to get interesting/useful Maths.
@nprofile…cqfe @nprofile…tczy

2025-12-26 18:36:04 UTC

- reply

@nprofile…8eua There are infinite cool things to learn and a finite time to learn them. You have to prioritise 😉

2025-12-26 12:31:18 UTC

- reply

@nprofile…8eua I didn't mean to offend you. If I did, I apologize.
I understand this stuff on an intellectual level, but I never devoted nearly enough time to music to make this stuff less confusing at an intuitive level. There is just too much historical baggage.

2025-12-26 12:16:53 UTC

- reply

@nprofile…8eua The very fact that people decided to divide notes into "white" and "black" instead of simply say that they are ultimately the same thing is also something that will confuse me forever 🤣

2025-12-26 12:02:31 UTC

The fact that going up 7 notes is called a fifth will never cease to confuse the hell out of me.
#IAmBadAtMusic
#nevent1q…l9me

2025-12-19 11:05:23 UTC

Reminder: "This is a bubble" and "This has no practical applications" are two very different statements.

2025-12-18 11:36:55 UTC

- reply

@nprofile…8eua After not touching it for >1y, I used this occasion to deactivate it (it will take 30 days before it is deleted).

2025-12-15 14:14:02 UTC

If you are not following Bruno, and you have any interest in Physics / Computers / Computer Graphics / Computational Physics &co, do yourself a favour and start following him.
#nevent1q…cex9

2025-12-15 12:44:32 UTC

- reply

@nprofile…8eua https://www.unicodeit.net/ has saved me so many times.
@nprofile…p3nn

2025-12-05 09:03:08 UTC

To be fair, in German you could build the word "freudenfreude", but I agree that "mudita" sounds better and is easier to spell 😉
#nevent1q…yqdl

2025-12-02 14:40:17 UTC

Stuff like this is why I hate generative AI with a passion.
https://www.theguardian.com/technology/ng-interactive/2025/dec/02/jared-kaplan-artificial-intelligence-train-itself
We know for a fact that feeding a generative AI model with the output of generative AI lead to the model collapse and nothing more ( https://www.nature.com/articles/s41586-024-07566-y ), but the claim is not just factually incorrect, it is maliciously incorrect. Because they know it is false, but they will pretend otherwise to try to keep the speculative bubble afloat and the money flowing in.
And the whole article is nothing else but the "journalist" repeating the claims of someone with a very clear vested interest without ever fact-checking or even questioning them.
It is rage-inducing.
#AI #GenerativeAI #Anthropic #rant

2025-11-27 07:57:27 UTC

- reply

@nprofile…pklh If I split it into 3 parts, can I then recombine them to obtain 2 spheres identical to the original?

2025-11-11 15:36:17 UTC

Need to prepare a lecture on applications of Monte Carlo integration, and I was wondering if to talk 20-30 min about path/ray tracing.
Problem: I am a Physicist, and all this computer graphics for me has always been "and here they numerically approximate the radiative transfer equation" and I never looked into how to implement any of that.
Anyone has a Physicist-friendly resource I can look at? 🙏
(Maybe @nprofile…d23k @nprofile…cpuy @nprofile…nuqs ?)
#ITeachPhysics #ComputerGraphics #MonteCarlo

2025-11-06 15:33:19 UTC

There is no shortage of automation games nowadays (from Factorio onward) and I think there are a couple of good SimCity-like games out there?
But has anyone dared to make something like Sim-Life, but with the power of modern computers to simulate planet-scale ecologies?

2025-11-05 11:50:11 UTC

- reply

@nprofile…tqek Danno assolutamente per scontato che sia cattolico. L'ipotesi che un bianco possa seguire una qualunque altra religione (o, orrore degli orrori, essere ateo) non li sfiora neppure
@nprofile…tzrf

2025-11-04 10:35:16 UTC

- reply

@nprofile…8eua I wonder if Meitner ever crossed paths with Noether. Both Jews women in Germany struggling to have their genius recognized by a close-minded University system. Both having to flee Germany. Roughly the same age.
They were interested in very different problems, but I can only imagine what they could have achieved if they ever worked together!

2025-10-28 14:09:37 UTC

ChatGPT costs more money and resources to run than Wikipedia. So if the answers from ChatGPT are not at least as good as the ones you can find just visiting the correct Wikipedia page, the rational move is to give all of OpenAI money to the Wikimedia foundation.

2025-10-18 10:56:35 UTC

- reply

@nprofile…4stm Knitting is the kind of activity that disturbs nobody, so it would be very petty to forbid it 🙂

2025-10-18 07:40:14 UTC

There is a girl who started brining her knitting to my class, and now I wish I could spend a few lectures on knot theory and (some of) the beautiful Maths behind knitting 🧶

2025-10-16 09:59:22 UTC

When you see people screaming that the "establishment" is suppressing their their discoveries / insights / theories / etc they have invariably taken the path on the right.
99.99% of the scientific community will squeak in joy if you ever came up with something that truly challenged the currently accepted theories, but we have exhausted our patience with BS and cranks a long time ago.
#Science #Physics #PseudoScience #CrankScience #xkcd
https://xkcd.com/3155/

2025-10-15 08:45:29 UTC

- reply

@nprofile…ah9w "For instance, someone might embed such triggers in specific websites to make models unusable when they retrieve content from those sites."
I now need instructions on how to include the poisoned text in my website so it is invisible to a human reader (e.g. as tiny white text? Would that work?)

2025-10-13 10:49:52 UTC

#PhysicsFactlet
Sometimes you need to forgo the intuitive way to define stuff for the sake of actually being able to do anything useful with those definitions.
An example of this I always found funny is in knot theory, where a simple loop is considered to be a knot, while anything where the two extremes are dangling are not, including the common overhand knot.
This looks weird the first time you see it, but there is a very good reason to go with such a definition: you want to study what you can and can't do by manipulating the knot, and if you have the two extremes dangling, you can always untie any knot, making them all equivalent to a piece of string. In order to be able to say anything interesting about them you need to remove this trivial option, and thus accept the simple loop as a knot.
#knots
https://media.mathstodon.xyz/media_attachments/files/115/366/428/534/174/821/original/0c89383034a0cc2c.png

2025-10-08 10:14:53 UTC

- reply

@nprofile…k63y Thermodynamics feels called out 🤣

2025-10-07 16:50:50 UTC

- reply

@nprofile…fa4k A "Clifford algebra for dummies" would be nice.
(I am the dummy.)

2025-10-03 11:13:44 UTC

#PhysicsFactlet
A geodesics on a surface is the shortest* curve connecting two points.
*or longest, if you want to be pedantic 😉
https://media.mathstodon.xyz/media_attachments/files/115/309/934/107/591/247/original/ed06bc39d666a81d.mp4

2025-06-10 08:41:34 UTC

- reply

@nprofile…hr6z I genuinely hope there are, but I doubt it. Here they started going full speed toward MS ~10 years ago and they have yet to slow down.

2025-06-06 13:39:30 UTC

- reply

Look mom, I got a semi-viral post on Mastodon!
I don't have a soundcloud (or whatever people who regularly go viral tend to advertise), so just be nice to each other!

2025-06-06 08:50:06 UTC

"AI companies claim their tools couldn't exist without training on copyrighted material. It turns out, they could — it's just really hard. To prove it, AI researchers trained a new model that's less powerful but much more ethical. That's because the LLM's dataset uses only public domain and openly licensed material."
tl;dr: If you use public domain data (i.e. you don't steal from authors and creators) you can train a LLM just as good as what was cutting edge a couple of years ago. What makes it difficult is curating the data, but once the data has been curated once, in principle everyone can use it without having to go through the painful part.
So the whole "we have to violate copyright and steal intellectual property" is (as everybody already knew) total BS.
https://www.engadget.com/ai/it-turns-out-you-can-train-ai-models-without-copyrighted-material-174016619.html?src=rss

2025-05-31 09:01:07 UTC

- reply

@nprofile…m2ru @nprofile…f7v6
Not if you think about it. If you can imprecisely control a lot of phasors, you can make the point in roughly the same direction. They won't be perfectly on a line, but on average they will still interfere constructivly. Controlling a lot of them is more important than controlling them perfectly.

2025-05-30 23:50:09 UTC

- reply

@nprofile…m2ru @nprofile…f7v6
It depends a lot on exactly what you want to do, but to create a simple focus like in this visualization you can get surprisingly good result with very poor precision, as long as you are stable and you can control a large number of degrees of freedom.

2025-05-30 10:37:43 UTC

#PhysicsFactlet
Scattering scrambles coherent light into a speckle pattern, where the field at each point can be seen as the superposition of a large number of random phasors. At some point the result is brighter, and at some points the result is dimmer, creating the "speckly" pattern.
By changing the phase of the incident light one can change the phase of the phasors making up the resulting field, and since elastic scattering is linear, changing the phase of different input modes is going to rotate different phasors without cross-talk.
As a result it is possible to find an incident wavefront such that all the phasors making up the field at one point are in a straight line (constructive interference), resulting in a single bright dot (a focus) through a completely scattering material.
#Optics #Physics
https://media.mathstodon.xyz/media_attachments/files/114/596/310/441/654/962/original/0241c23b0d5bf003.mp4

2025-05-23 15:54:49 UTC

#PhysicsFactlet
A Shack-Hartmann sensor is made my an array of small lenses and a camera. If the light hitting the lenses is collimated, we will get a number of equispaced foci on the camera. But if the light is not collimated, the position of the foci will change in a predictable way, so we can reconstruct where the ray were coming from.
#Optics #Physics
https://media.mathstodon.xyz/media_attachments/files/114/557/948/986/672/808/original/f79c4870825557b5.mp4

2025-05-19 08:45:33 UTC

#PhysicsFactlet
Inspired by a short thread by @nprofile…fa4k (https://mathstodon.xyz/@johncarlosbaez/114459727454034873) I made a visualization of a Stone-Wales defect in fullerene, released it into the #PublicDomain, and uploaded it on Wikimedia Commons.
https://commons.wikimedia.org/wiki/File:Stone-Wales_defect_in_Fullerene.gif

2025-05-19 08:34:24 UTC

For the people at Wikimedia Commons: I used to be User:Berto, but I lost my password, so I now I am User:Clodovendro. I make a number of scientific visualizations using #Mathematica, and I release them into the #PublicDomain (and upload the on Commons so everyone can use them).

2025-05-12 16:31:58 UTC

- reply

@nprofile…fa4k Travelling tomorrow, but will upload it and link it as soon as I find a bit of time 🙂

2025-05-12 16:15:20 UTC

It looks like it is #webcomicday, so here are some of the webcomics I read regularly (in alphabetical order).
* Aurora (https://comicaurora.com/). A high-fantasy adventure involving a lot of minor gods and gods' avatars. From Red of Overly Sarcastic Production fame.
* Fantasy High (https://www.webtoons.com/en/comedy/fantasy-high/ep-1-the-beginning-begins/viewer?title_no=7686&episode_no=1). A webcomic adaptation of the homonymous show by #dimension20.
* Girl Genius (https://www.girlgeniusonline.com/). THE steampunk saga, from Phil and Kaja Foglio.
* How to be a dragon (https://www.webtoons.com/en/canvas/how-to-be-a-dragon/dragon-breath/viewer?title_no=696410&episode_no=1). A cute little webcomic about being a little dragon.
* Kill Six Billion Demons (https://killsixbilliondemons.com/). One of the best things I have read in ages. An adventure of truly cosmic scale.
* Magefront (https://www.magefrontcomic.com/comic). A necromancer in training for the military, her wraith, and the friends she made while trying not to think about the ongoing war.
* The Order of the Stick (https://www.giantitp.com/Comics.html). A classic among classics. What if the world, but DnD?
* Saturday Morning Breakfast Cereal (https://www.smbc-comics.com/). Some of the smartest humour you can find on internet.
* Sellswords & Sorcery (https://www.webtoons.com/en/comedy/sellswords-and-sorcery/list?title_no=6828). An unlikely band of mercenary.
* Side Quested (https://sidequested.com/). Is this girl going to save the princess? Yes. And then some.
* STARS (https://www.webtoons.com/en/canvas/stars-the-webcomic/list?title_no=743991). Star-trek-style, but a lot more surreal.
* Swords (https://swordscomic.com/comic/cover/). What if the world, but swords?
* Ten Earth Shattering Blows (https://tenearthshatteringblows.com/archives/comic/1x01). Sort of fantasy with a Mad Max vibe. By @nprofile…mjsg * The Weekly Roll (https://www.webtoons.com/en/canvas/the-weekly-roll/list?title_no=358889). Are we the baddies? Yes, but we are trying not to be.
* XKCD (https://xkcd.com/). If you don't know XKCD, what are you even doing?

2025-05-12 14:51:50 UTC

- reply

@nprofile…fa4k Third time's a charm? 🙂
@nprofile…4ynp
https://media.mathstodon.xyz/media_attachments/files/114/495/420/743/377/392/original/98a410848799a586.mp4

2025-05-12 12:42:36 UTC

- reply

@nprofile…fa4k Would it make sense to always keep the red part in front and make it rotate only 180 degrees forth and back?

2025-05-12 10:40:00 UTC

- reply

@nprofile…fa4k Any better?
https://media.mathstodon.xyz/media_attachments/files/114/494/429/577/960/394/original/7f06130c5bef197c.mp4

2025-05-11 11:07:56 UTC

- reply

@nprofile…fa4k Shared on bsky with my "active particles" expert (Giovanni Volpe) to see if he has an understandable explanation.

2025-05-09 09:40:19 UTC

- reply

@nprofile…fa4k draft number 1. Feedback is welcome 🙂
(I know the animation will benefit from being smoother, but I won't generate a ton of frames for a draft)
https://media.mathstodon.xyz/media_attachments/files/114/477/201/989/778/967/original/032bbb7157376e68.mp4

2025-05-08 08:53:35 UTC

- reply

@nprofile…pac7 : is blue noise hyperuniform? (I think it is, but you are the expert)
@nprofile…fa4k

2025-05-06 07:55:54 UTC

- reply

@nprofile…fa4k If I find a bit of time I might take up the challenge 😁

2025-05-05 10:39:43 UTC

- reply

@nprofile…fa4k Depressing but not surprising.
(Deleting anything that smells of crackpottery on sight is a strategy that has served me well through the years)

2025-05-04 09:05:54 UTC

- reply

@nprofile…fa4k They should make him a honorary retired engineer 🤭

2025-04-26 09:06:59 UTC

- reply

@nprofile…mmm2 No European country can absorb even a tiny slice of US academia (nor can Canada, Japan, Korea etc). We already have way more smart and hard working people wishing for an academic career than places in academia.

2025-04-25 08:59:57 UTC

"Perplexity is building a browser in part to collect customer data for targeted ads"
Does anybody knows where I can but some torches and pitchforks?
https://www.engadget.com/ai/perplexity-is-building-a-browser-in-part-to-collect-customer-data-for-targeted-ads-230132091.html?src=rss

2025-04-23 21:01:10 UTC

- reply

@nprofile…fa4k
@nprofile…kppt might have an opinion

2025-04-23 14:27:32 UTC

#PhysicsJournalClub
"Temperature as joules per bit"
by C.A. Bédard, S. Berthelette, X. Coiteux-Roy, and S. Wolf
Am. J. Phys. 93, 390 (2025)
https://doi.org/10.1119/5.0198820
Entropy is an important but largely misunderstood quantity. A lot of this confusion arise from its original formulation within the framework of Thermodynamics. Looking at it from a microscopic point of view (i.e. approaching it as a Statistical Mechanics problem) makes it a lot more digestible, but its ties to Thermodynamics still creates a lot of unnecessary complications.
In this paper the authors suggest that by removing the forced connection between entropy and the Kelvin temperature scale, one can rethink entropy purely in terms of information capacity of a Physical system, which takes away a lot of the difficulties usually plaguing the understanding of what entropy is actually about.
I don't think the SI will ever consider their suggestion to remove Kelvins as a fundamental unit and include bits, but this paper will be a great boon to any student banging their head against the idea of entropy for the first (or second, or third) time.
#Physics #Entropy #Thermodynamics #StatisticalMechanics

2025-04-17 14:44:36 UTC

- reply

@nprofile…e5g0 If you give me all the money in the world I promise to create an orchestra of 10⁵ elements, able to produce the best music in the world!
(tl;dr: it's a grift)

2025-04-09 14:49:34 UTC

#PhysicsJournalClub
"Direct observation of colloidal quasicrystallization"
by Y. Gao, B. Sprinkle, DWM Marr, and N. Wu
Nat. Phys. (2025)
doi.org/10.1038/s41567-025-02859-z
Quasicrystals are weird. When you solidify something it tends to get into a high-order state: a crystal. If you cool it down too fast so it doesn't manage to make a monocrystal it will form a polycrystalline state or, worse case scenario, something completely amorphous like a glass.
But quasicrystals are weird. They are ordered structures that lack a periodicity and making them is not easy.
In this paper the authors show how paramagnetic colloidal microspheres (i.e. big enough to be clearly visible under a microscope) subject to an electromagnetic field spontaneously arrange themselves into a quasicrystal.
This is 100% not my field, but the ability to create quasicrystals on demand looks so cool!
#Physics #Chemistry #Quasicrystal

2025-03-29 09:11:05 UTC

- reply

@nprofile…fdgy
A lot of Physics high-performance code is still written in Fortran. Apparently you just can't beat how good the compilers are.
@nprofile…mzyx @nprofile…vzvd

2025-03-20 07:50:57 UTC

- reply

@nprofile…yu7x One more reason to move ALL scientific conferences out of the US.

2025-03-12 07:39:43 UTC

- reply

@nprofile…fa4k @nprofile…m7f4 Anything bigger than a standard Russel's teapot is a moon 🙃

2025-02-27 16:01:20 UTC

- reply

@nprofile…3ug8 Immagino che tu conosca già il "password game", ma giusto nel caso te lo metto qui che magari ti serve per fare allenamento 😉
https://neal.fun/password-game/

2025-02-27 09:36:31 UTC

I would expect all my fellow scientists to be able to spot a dishonest chart from 10 miles away, but it is still good to get a quick refresher from time to time.
Also a good resource to give to students, who are often very naive on these things.
https://flowingdata.com/projects/dishonest-charts/

2025-02-05 09:51:38 UTC

#PhysicsJournalClub
"The Tumultuous Birth of Quantum Mechanics"
by Philip Ball
https://physics.aps.org/articles/v18/24
"quantum mechanics wasn’t created all at once. It took several decades and was a messy, confused process, during most of which the true nature of this revolution was obscure. In some ways it still is."
#QuantumMechanics #HistoryOfScience #Physics

2025-01-25 18:20:48 UTC

Feels like the right time to start planning for moving all international scientific conferences to somewhere outside the US.

2025-01-24 11:10:38 UTC

- reply

@nprofile…yu7x
https://media.mathstodon.xyz/media_attachments/files/113/883/018/208/883/008/original/80b43cc30b979be6.png

2025-01-13 16:35:33 UTC

- reply

@nprofile…dp56 I tried both, but couldn't come up with a simple formula where the frequency was position-dependent, the line was nice (continuous and infinitely differentiable) and didn't take forever to compute 😉

2025-01-13 09:45:35 UTC

- reply

@nprofile…dp56 This looks a LOT better than my rudimentary "sinusoidal Mario". 😃
Do you have a legible code I can take a peek at for inspiration?
https://media.mathstodon.xyz/media_attachments/files/113/820/394/283/285/852/original/b7e0a8d340328ab1.mp4

2025-01-13 07:45:27 UTC

In a rational world insurance companies would be at the frontline, lobbying for pro-environmental policies to minimise the number of people dying, getting sick or having their house burnt down.Just to avoid having to pay.
And yet...

2025-01-09 11:21:49 UTC

#PhysicsJournalClub
"Blending Optimal Control and Biologically Plausible Learning for Noise-Robust Physical Neural Networks"
by S. Sunada et. al.
Phys. Rev. Lett. 134, 017301 (2025)
I am not a big fan of generative AI, especially in the form it is sold and advertised today, bvut I think that there is a lot of cool and potentially useful research been done in the field of Machine Learning. In particular I am very interested in approaches where most of the heavy lifting is left to Physics, i.e. the Physical system naturally does stuff, and you exploit it for Machine learning, instead of spending a ton of resources and energy to create and manipulate the whole neural network.
In this paper the authors borrow ideas from different sub-fields to put together a training method that doesn't really need to know much about what the Physical system is actually doing, and doesn't need an accurate control of the Physical system either.
I admit I understand no more than 10% of what they say (Machine Learning is definitively not my field), but it looks interesting and promising!
https://doi.org/10.1103/PhysRevLett.134.017301
#MachineLearning #Physics

2025-01-07 17:20:42 UTC

- reply

@nprofile…36ju This is not very different from what any EU citizen has to do to enter the US for a short visit.

2025-01-07 10:17:47 UTC

#PhysicsJournalClub
"Testing the necessity of complex numbers in traditional quantum theory with quantum computers"
by Jarrett L. Lancaster and Nicholas M. Palladino
Am. J. Phys. 93, 110 (2025)
In classical electrodynamics the use of complex numbers is only due to to its convenience for calculations. Nobody wants to remember all those pesky trigonometric identities, so we use complex numbers to simplify calculations and take the real part at the end. You need to be a bit careful when calculating stuff like the Poynting vector, but this is well addressed ion any half-decent undergrad-level textbook.
But for quantum mechanics the problem is less obvious. On one hand we only ever measure real quantities, but on the other hand the imaginary unit appears explicitly in the Schrödinger equation, and no textbook I am aware of ever even mention the possibility that quantum numbers might be just a calculation convenience like it is in classical electrodynamics.
The question is subtle enough that you are going to find no shortage of well-read Physicists claiming that it is "obvious" that complex numbers are necessary for quantum mechanics, or that it is "obvious" that you could just use real numbers if you wanted.
This paper makes a pretty good job at explaining the problem, going through some of the history and explicit calculations, up to constructing explicitly a real-valued version of QM.
The second part, where they make an "experiment" on a IBM cloud quantum computer is (imho) less interesting, and their conclusion that you need indeed complex numbers not really supported by the evidence, but your mileage might vary 🙂
https://pubs.aip.org/aapt/ajp/article/93/1/110/3327097/Testing-the-necessity-of-complex-numbers-in
#Physics #QuantumMechanics #ITeachPhysics

2025-01-06 19:21:26 UTC

- reply

@nprofile…fa4k I am missing the experimental evidence of black holes somewhere in that list. Not sure where one would put the exact date, but when I was a student it was still not obvious at all that black holes were real 🙂

2025-01-06 12:44:03 UTC

#PhysicsFactlet
Schematic of the epicycle model, as requested by @nprofile…fa4k (see https://mathstodon.xyz/@johncarlosbaez/113652794327730499 for his thread on the topic).
Apart from having (hopefully) more colorblind-friendly colors, this one is in the #PublicDomain instead of being copyrighted.
#Physics #Astronomy
https://media.mathstodon.xyz/media_attachments/files/113/781/438/824/005/821/original/f1bd725d46a98b4c.mp4

2025-01-01 12:54:07 UTC

- reply

@nprofile…yu7x We finally got the killer application for LLMs! 🥳

2025-01-01 12:25:11 UTC

- reply

@nprofile…yu7x What I am reading is that there is a small but nonzero chance of every day being Saturday/Sunday!

2024-12-31 20:51:13 UTC

Yearly repost.
Happy new year!
https://media.mathstodon.xyz/media_attachments/files/113/749/384/904/279/638/original/b0fab2a802e22787.png

2024-12-21 15:21:31 UTC

I recently started replaying Katamari Damacy (my 6y/o finds it hilarious) and I have to say I didn't remember the soundtrack to be so good!
(I don't understand a single word of Japanese, so I have no clue what the lyrics say, and it is probably better this way.)
#videogames #katamari

2024-12-13 08:15:43 UTC

Does anybody knows where I can find reliable numbers on how mouch energy (and/or CO2 equivalent) a generative AI like chatgpt consumes per prompt?

2024-12-07 18:57:02 UTC

- reply

@nprofile…yu7x The "science" feed is largely pointless. The "Physics" feed is marginally better. Still, nothing beats curating who you are following and looking at the feeda once in a while to check if there is someone there worth following you didn't know about.

2024-12-02 09:59:07 UTC

Scientist types as #DnD classes:
Fighter: The best way to approach any problem is to do an experiment first and ask questions later (assuming there is any question left to ask). Very high volume of fire, but needs a constant stream of money to pay for fancier and fancier equipment.
Wizard: You dive deep into the mathematical foundations of your discipline. So deep you lost contact with reality a long time ago, but it doesn't matter, as you only speak with other wizards, who also lost contact with reality a long time ago.
Cleric: Never the first nor the last author. Your name is always somewhere in the middle, like an afterthought. But it is to you that the rest of the party always comes back to for help every single time they get in troubles.
Druid: You really care about the impact you make, and find the mathematical formalism at the foundation of your discipline a waste of time. So you developed a more intuitive approach, and rely a lot on your gut feelings.
Sorcerer: You are naturally gifted, and thus you never had to really study to be successful at what you do. You are a raising star, but true wizards look down on you as a shallow amateur (while at the same time envying you).
Thief: You are somewhere. Everybody knows that. But no one seems to be able to say where or doing what. Officially you are working for the good of the collaboration, and since everything seems to be working out everyone assumes you are doing your part.
Warlock: You are not as gifted as a Sorcerer, nor you put as much effort in as a Wizard. What you have is an extremely powerful patron, who grants you visibility, and the money you need to carry out your research. All for the small price of your soul.

2024-12-01 21:45:43 UTC

- reply

@nprofile…fa4k I assume you know @nprofile…kppt and his blog (in particular the series "From τὰ φυσικά to physics": https://thonyc.wordpress.com/from-%cf%84%e1%bd%b0-%cf%86%cf%85%cf%83%ce%b9%ce%ba%ce%ac-ta-physika-to-physics/ )

2024-11-27 17:29:27 UTC

- reply

@nprofile…ucsf
Any hack that works will be, sooner or later, formalized, given a name and rise to "method" status.
@nprofile…j07c @nprofile…dpz7

2024-11-27 11:10:51 UTC

- reply

@nprofile…yu7x That smells of sunk cost fallacy 😜

2024-11-27 10:53:29 UTC

- reply

@nprofile…yu7x Not your point, but t a certain point you will stop seeing them, as all the light they reflect/emit will be redshifted out of the visible range 🙂

2024-11-26 20:13:57 UTC

"Zoom is now an AI-first work platform"
Does anybody know of a half-decent videocall app that is not infested by AI? I need one.
https://www.theverge.com/2024/11/25/24305942/zoom-communications-rename-ai-first-company

2024-11-24 16:12:17 UTC

- reply

@nprofile…yu7x @nprofile…e5g0 @nprofile…fa4k @nprofile…cnse @nprofile…j7pr
This whole thread is a beautiful representation of the best social media have to offer in terms of scientific discussion.
Thank you to all the people involved, I love it! 🤩

2024-11-22 10:08:08 UTC

- reply

@nprofile…e5g0
As far as I understand GR, even long before you actually arrive at the proper horizon things will get very dark, as only the light moving in a narrower and narrower cone will be able to escape.
@nprofile…fa4k @nprofile…j7pr

2024-11-14 15:30:54 UTC

- reply

@nprofile…vrgz To be on the safe side, send the homework to the prof together with the screenshot, so you can truthfully claim you did everything you could.

2024-11-14 07:59:44 UTC

- reply

@nprofile…gx3z
Raccontamela meglio questa, che non la conosco.
@nprofile…v78a

2024-11-06 15:45:38 UTC

- reply

@npub19qm…sn8k A felony is only a felony if you are poor.

2024-11-06 07:58:34 UTC

One day we will have to face the fact that a whole lot of people actually really like the idea of fascism.
And no, I have no idea if anything can be done about it.