The h Bar

John Rennie

17:00

Windows Subsystem for Linux (WSL) is a compatibility layer for running Linux binary executables (in ELF format) natively on Windows 10. WSL provides a Linux-compatible kernel interface developed by Microsoft (containing no Linux kernel code), which can then run a GNU userland on top of it, such as that of Ubuntu, openSUSE, SUSE Linux Enterprise Server, Debian and Kali Linux. Such a userland might contain a Bash shell and command language, with native GNU/Linux command-line tools (sed, awk, etc.) and programming language interpreters (Ruby, Python, etc.). When introduced with the Anniversary Update...

I've used it and it works really well. However I'm not sure about CUDA ...

enumaris

bitcrazed
6 points
·
3 months ago
Yes, CUDA support is on our radar, but understand that adding GPU compute support is complex and will take some time & effort to deliver.

this was posted 3 months ago in a reddit thread

probably not supported yet if I am to guess...3 months is pretty short

Emilio Pisanty

@ConstantineBlack if you're thinking about staying in academia, then sure, go for it

enumaris

I've never used something like that though so I dunno much about it...

John Rennie

No, it appears WSL does not support CUDA. Oh well, dual boot it is then.

Emilio Pisanty

if you suspect you might want to leave, then a theoretical mathematics masters deserves a bit more consideration than other, more applied, options

enumaris

17:04

I would need them to take a lot of this junk out of my SSHD

Semiclassical

hmmmm

enumaris

they filled my SSHD to the brink with preloaded crap...-.-

I wouldn't have enough space for a real dual-boot

Emilio Pisanty

i.e. you should consider carefully what types of transferable skills you should develop with, through, and alongside your masters, which will be more broadly useful if you decide to leave

Semiclassical

one nice property of the two-electron singlet state is that it has angular momentum zero and therefore must have rotational symmetry

Constantine Black

@EmilioPisanty Like something with coding, software developing and such, or you mean other areas of physics too?

John Rennie

17:05

@ConstantineBlack ping @Danu here

Emilio Pisanty

@ConstantineBlack exactly

Semiclassical

for pure states, I think that's the only such two-qubit state with that property

Is that also true for mixed states? I feel like it should be

Emilio Pisanty

i.e. given two equivalent academic projects that you're equally interested in, choose the one that forces you to learn technology stacks that are more broadly applicable

enumaris

pain

Constantine Black

@Danu Greetings. Is there a chance you could give some info on this: Could I ask for opinions, if anyone cares to give, about a possible transition from a physics bachelor to a theoretical mathematics(algebra, diff. geo.) master? Any thoughts are welcomed.
@Danu Any possible opinion is welcomed.
@EmilioPisanty From that perspective, mathematics should not be a priority. Maybe search some coding plus theoretical physics project( nuclear physics, lattice qcd) and learn mathematics for that reason?

danielunderwood

17:09

Since you guys are talking about academics, any advice for someone that was a terrible undergrad and may be interested in going back to school? I have a good recommendation from my research advisor, but was pretty lazy in classes

enumaris

going back to school as in you dropped out of undergrad or you mean like going to a grad program?

danielunderwood

Like grad program

Semiclassical

how long ago did you graduate?

danielunderwood

2 years ago

Semiclassical

hmm

enumaris

17:11

Many grad programs are pretty heavily research based - if you have good research to show and a good direction, I think that goes a long way towards acceptance.

having a prof who will advocate for you goes a long way as well

John Rennie

@danielunderwood Doing a PhD is pretty hard. You need to be well motivated. Are you sure you have that much commitment?

Emilio Pisanty

@ConstantineBlack that's not what I said.

> given two equivalent academic projects that you're equally interested in

... but of course, it feeds back to my original question, what are you going to do with the degree?

I'm not saying "do not enroll for a degree that provides thin job prospects outside of academia"

I am saying "do not enroll for a thin-job-prospects degree without due consideration that that is the case"

Semiclassical

walk in with your eyes open

Emilio Pisanty

If you really like the material and you intend to stay in academia then by all means go for it

Semiclassical

and with an eye towards with what you actually want to get out of grad school

Emilio Pisanty

17:14

if not, then I don't like to be the bearer of bad news, but it can be a recipe for some pretty tough years during and after grad school

@rob is it actually small? i.e. do we know that for a fact?

1

Q: Is it possible to estimate the active reviewer population using the SEDE data?

My main site has, I think, a problem with a shrinking reviewer population, and I would like to use the SEDE data to estimate whether this is indeed the case and, if so, how bad it is and how it has evolved over time. As such, I would like to use the SEDE data to answer specific questions of the ...

^ please consider supporting

danielunderwood

@JohnRennie Yeah I believe so. I'd ideally like to do some non-degree courses or a masters first with the option to continue if I do well just to make sure. I know that isn't really the typical path though

John Rennie

Doing a masters is a pretty standard way to get into doing a PhD. Many PhD programmes in effect do a masters as a first year and you have to pass it to continue.

danielunderwood

Ahh I was under the impression that people just normally were enrolled in a phd program from the start and picked up a masters along the way

Constantine Black

@EmilioPisanty Yes, I know. I was meant to put a ? there. And I see your case; to rephrase, if I had the possiility of choosing between a math grad and a computational physics grad, I am closer to the first at this time, although I will have to work more on some technical knowledge like coding.
And I appreciate your consideration, you are not saying something I haven't already thought- so I mean it's not a bad news. Let me put it this way: Could a mathematics master help in pursuing theoretical physics or then I will have to pursue another master in theoretical physics( if such a question m…

Semiclassical

@danielunderwood over at my uni that's how it is. but typically if you don't pass the written exam on your first or second try, then you get put on the masters track and have to do a bit more legwork in that regard

I think the point is to what extent getting the 'masters first' is formalized.

John Rennie

17:19

@danielunderwood I really, really, REALLY, enjoyed doing my PhD, but I was a major hard core nerd and couldn't imagine living any other way. If you've been working for two years, and presumably earning real money, then going back to the impoverished life of a phD student is going to be a wrench.

3

Emilio Pisanty

@ConstantineBlack it depends on the track you want to take

but yes, it can definitely help

Semiclassical

External factors also matter (ofc)

Emilio Pisanty

But if that's the question, why are you asking us? Go find a bunch of PhD programmes that you'd be interested in applying for (some five-ten), find out who runs them, and email them with the question.

Semiclassical

e.g. if you've got dependents you're supporting, or student loan debt you'll be needing to pay off

Emilio Pisanty

@danielunderwood that completely depends on where you are

dmckee

17:21

The people I know personally who did the "go back to school for an advanced degree after working for a while" fall into two categories.

Emilio Pisanty

91

A: Welcome to Academia.SE!

Academia varies more than you think it does Academic customs and procedures vary greatly across countries, universities, fields, subfields, workgroups and so on. Therefore always consider that what you assume to be general in your question or answer is not. It is very helpful if your question i...

dmckee

(A) Want a degree (mostly a masters) for specific job-advancement reasons.

Emilio Pisanty

@dmckee : to "?

dmckee

(B) Well into their career, have their lives well under control (mostly empty nesters at that), and do it part-time because they passed on the chance for practical reasons when they were younger.

Constantine Black

@EmilioPisanty I already have began talking with people( mathematics and physics)- money is an issue. Just asking here because someone could have some experience. Thanks for your thoughts :).

dmckee

17:23

I did have a colleague in grad-school who was retired from a government career.

@EmilioPisanty Tempted to write a mini-language which uses that syntax for something. Just because it would be weird.

Only I'm noodling a iOS app right now, and learning a bit about swift along the way.

I should say that the people I've known in group (B) are generally quite satisfied with the life they have had and don't regret their earlier choice.

It's just that now they can chase that other dream a little, too.

Anonymous

@dmckee How's the performance of such candidates compared to the other students in the program, usually in terms of enthusiasm, work-ethic, grasping abilities etc. ? Does staying away from school for long make it hard to get back into studies (drastically)?

dmckee

@Blue In the cases where I know in detail they had to do a little recovery work on their math before really getting started, and they complained about not being as fast as they were when they were younger.

danielunderwood

@dmckee I don't know how much I fit into either of those. I'm not terribly worried about a degree for career advancement for the sake of career advancement. It's more of wanting to follow my curiosity and wanting to return to the research world. Don't really know how great of a reason that is though. I certainly don't fall into the second group, though I also don't have a family that I have to worry about supporting or influencing my decisions at the moment

dmckee

But they also generally have better work habits and time management, which ends up putting them in the middle of the pack.

Semiclassical

So, looping back to the states question I had earlier

Anonymous

17:31

Makes sense. I guess the more technical the degree is, the harder it is to get back

Anonymous

@dmckee That's a good point

dmckee

@danielunderwood The dependents are the hard part. Unless your spouse is able and willing to do a big part of the bread winning for a while.

Graduate stipends in physics and math are quite generous by the standards of the humanities (I don't know how anyone does an advanced degree in those fields), but they are hardly plush.

Semiclassical

being able to live with your parents is another route, though that comes with other problems (and is hardly an option in many cases)

student loan debt is another big one

vzn

@Semiclassical (ps) its a list. how about you pick any one that interests you. they are all credible. the theory will take work to push fwd. its not yet servable on a silver platter so to speak. as always am willing to dedicate time to serious/ genuine co-working on it. also consider actually responding to my repeated invitations. feel a kind of love/ hate jekyll-hyde dynamic at moment am sure you can realize. :( o_O

Anonymous

@Semiclassical Having to listen to their orders? :P That's a thing

Semiclassical

17:35

@vzn dude, signal-to-freaking-noise

vzn

oops hyde mode wrong timing my bad :(

Semiclassical

giving someone a list of dozens of possible refs and saying "they're all credible, pick one" is not clarifying

vzn

@Semiclassical so you do want it served on a silver platter. which it is :P ps am very tired of ppl dismissing very credible refs as not credible without even looking at any of them.

danielunderwood

@dmckee Yeah I've been lucky enough to save up and work on loans from working to help out if I do go back, though it likely wouldn't be much of an improvement to a stipend. On the bright side, I wouldn't be in as bad of a situation as a friend that went to med school

Semiclassical

then don't pretend you're answering my question.

vzn

17:38

ok. maybe its you who are repeatedly pretending. bye!

Semiclassical

back to actual physics:

enumaris

for reals u make much more in industry than academia generally speaking tho

Anonymous

@vzn Please don't expect people to read a dozen references which you took only a couple of seconds to cite, in order to verify whether what you're saying makes sense. JD does the same thing, and its annoying af.

Semiclassical

I tried the following: I generated an arbitrary 4-by-4 matrix $\rho$ (intended as the density matrix of the system). Defining $A(\vec{a})=(\vec{\sigma}\cdot\vec{a})\otimes I_2$ and $B(\vec{b})=I_2\otimes (\vec{\sigma}\cdot\vec{b})$, I imposed $\text{tr}(\rho)=1$, $\text{tr}(\rho A)=\text{tr}(\rho B)=0$, and $\text{tr}(\rho AB)=c(\vec{a}\cdot \vec{b})$. (the last expression is a sum of terms $a_j b_k$, so only the dot product has a hope of being rotationally invariant)

I then asked Mathematica to use SolveAlways to find for what matrix elements of $\rho$ those equalities would be true regardless of $\vec{a},\vec{b}$

And when the dust settled, I got the following 1D family (with basis kets $|S,m_s\rangle$

vzn

@Blue whatever. personally think ppl who endlessly pretend to be interested in (new) science are "annoying af." am not really asking anything of anyone in here (except maybe lining up words and actions?). that would be unrealistic expectations...

enumaris

17:42

haven't even gotten through all the old science

danielunderwood

@enumaris Yeah and they're a lot easier to get. I just don't really feel like I'm doing things that I find interesting most of the time.Makes a tough choice

bolbteppa

Haven't even understood the old science I got through so far

enumaris

my goal is to make enough money to do my own projects...

so far, I'm pretty far from this goal lol

danielunderwood

Like physics type projects or software?

bolbteppa

@vzn this fluid = everything stuff is not even science let alone new science

@Semiclassical why take tensor products of matrices and actually want their values :(

enumaris

17:47

Whatever I feel interested in...right now it'd be a mix between A.I. type stuff and Physics type stuff

Semiclassical

@bolbteppa because I want $\rho$ to be a rotationally invariant state

John Rennie

@vzn you keep saying bye, and raising our hopes, then you come back

enumaris

prolly more progress on the A.I. type stuff tho

@JohnRennie harsh bro, harsh

John Rennie

Harsh but fair.

bolbteppa

What about multiplying the actual values of the matrix $\boldsymbol{\sigma} \cdot \boldsymbol{a} = \begin{bmatrix} a_z & a_x - i a_y \\ a_x + i a_y & - a_z \end{bmatrix}$ with those of $\boldsymbol{\sigma} \cdot \boldsymbol{b}$

Semiclassical

17:49

@bolbteppa How do you think I'm getting Mathematica to do those products? :P

Anonymous

@vzn I feel you're already a step ahead from the general public since you're already quite interested in science. But I don't think simply being interested is enough. If you want people to take you seriously start with making solid technical statements when talking about science and cite specific and relevant sources rather than randomly citing anything and everything even vaguely related to the topic at hand. If you're not in the position to cite immediately, maybe read some more!

Semiclassical

(Mercifully, there's a KroneckerProduct command)

John Rennie

Anyhow, I urgently have to (a) fold my washing, (b) drink a cold beer and (c) finish my book. Back tomorrow at 5 a.m. as usual.

enumaris

beer is empty calories bro

vzn

@bolbteppa lol keep up the "encouragement" and maybe you can finally definitively convince semiclassical to give up bohmian mechanics which is inherently fluid mechanical (although maybe few realize this yet, even the supposed bohmian proponents).

John Rennie

17:51

@enumaris So is chocolate. And your point is? :-)

enumaris

well chocolate makes you feel full, beer doesn't

John Rennie

I don't eat chocolate in order to feel full.

I eat it for the pleasure I get from eating it.

enumaris

hmmm

alright then...proceed

John Rennie

@enumaris now I have your approval I can drink my beer with a clear conscience :-)

enumaris

:D

Semiclassical

17:54

$\rho(s) = \frac{s}{3}[|11\rangle\langle 11|+|10\rangle\langle 10|+|1\overline{1}\rangle\langle 1\overline{1}|]+(1-s)|00\rangle \langle 00|$

that's what I got as the most generic two-spin mixed state satisfying the above constraints

($\overline{1}=-1$, I just hate putting -1's inside kets)

where the kets are in the |S,m_s> basis not |m1,m2>

What I'm trying to convince myself of is whether that's actually a rotationally-invariant state

I don't think it can be but I'm not seeing an obvious counterexample

(also, $0\leq s\leq 1$ since $\rho$ is diagonal in this basis and therefore the diagonal elements are genuine probabilities)

I guess it really just boils down to whether $\rho(s=1)=\frac13\left[|11\rangle\langle 11|+|10\rangle\langle 10|+|1\overline{1}\rangle\langle 1\overline{1}|\right]$ is rotationally invariant

vzn

@Blue look agree you have a pt about JD being "annoying". dont want to turn into a door-to-door salesman of a worthless product. but have cited stuff endlessly on this site, typically ppl rarely read the refs even on request. am now concluding refs/ citations are not the main obstacle. there was a pt a few years ago where they were more scarce, but think a tipping point has been reached recently. feel not all the feedback is "in good faith"...

Semiclassical

hmm, maybe this is already known under the name "Werner state"...hmmmm

I guess that's consistent with what I wrote above, since the splitting is between a spin-symmetric part and a spin-antisymmetric part

enumaris

18:11

I need to find some good food for lunch...

maybe something with some hot soup...

Anonymous

@vzn Why not consider writing down the key points of those references in technical language (by that, I mean, using equations and theorems), so that people don't have to spend time reading references? I can't recall seeing any mathematical statement written down by you, ever. As an analogy, consider a product which has a ton of advertisement on the cover, but is hollow from inside.

Anonymous

The more you try to "advertise" non-mainstream theories without providing proof in technical words, it is natural that people will be repelled by you.

vzn

@Blue not all the criticism is constructive/ credible. perceive some major feinting going on not just by 1. have already done what you ask in my blog & in chat. let me know if/ when you read any of it. there is ample proof. there is some moving of the goalpost and scapegoating going on. people are repelled by new theories. am not taking it (all) personally.

Anonymous

@vzn And again, I don't see why you would expect people to spend time reading your blog posts without having any evidence first that you're capable of writing anything meaningful here

vzn

@Blue (speaking of feinting...) am taking you ~½ seriously. you havent given much sign you have any )( idea what the theory is about. all your words sound like completely sensible advice taken out of context.

Semiclassical

18:18

yep, i just rediscovered the two-qubit Werner state. okaly dokaly

enumaris

japan belgium is happening

vzn

← dont shoot the messenger™

Semiclassical

oh lol. of course. when $s=3/4$ then $\rho$ is just the identity matrix in the |S,ms> basis

vzn

@Blue since you raise it, another analogy to consider is a quality, even revolutionary new product that is lost in the noise... failing to gain traction due to misc factors eg lack of endorsements or whatever... some rather arbitrary...

bolbteppa

@vzn I bet if mainstream researchers began working on this stuff (i.e. if it made sense) you'd give up on them as too establishment pretty quick :p

2

vzn

18:23

@bolbteppa many/ numerous researchers are working on it. lack of understanding of that is (currently) a kind of tunnel vision or blind spot/ denigration/ devaluation. its a minority faction.

Semiclassical

tbf, that kind of a numbers game is not necessarily helpful in either direction

danielunderwood

I don't really know much about this argument and am not taking a side, but calling a specific line of thought revolutionary probably isn't the most credible way to present it

vzn

aka dismissal/ rejection/ ... stonewalling...

Cows

@Secret lol yes :P newscientist.com/article/…

Semiclassical

on the one hand, if the theory is correct then it doesn't matter if one guy is working or a thousand

vzn

18:28

lol theories do not advance by themselves.

Semiclassical

on the other hand, pointing to a collection of people working on a subject and saying "that's a lot of people" is not sensible to me either. if 20 people are working on it but there's only 200 people total doing similar research, then calling those 20 people "a lot" is a bit deceptive

vzn

this reminds me of the expr inside baseball and (not totally knocking it here) its amazing how much time can be wasted on it...

Semiclassical

that doesn't go to whether the theory is useful or not, but "a lot" is always a relative measure---"a lot" compared to what?

vzn

@bolbteppa this is reminding me of string theory too which is a major paradigm (gasp!) but even its rabid proponents eg LuMo admits that maybe only a few thousand in the entire world are familiar/ "working on it"...

Semiclassical

so while the number of researchers is not entirely irrelevant, it's a statistic which is liable to misuse by either proponents for or opponents of said subject

vzn

18:34

the # of ppl working on the theory is (to some degree) circumstantial evidence. exactly what pilot wave hydrodynamics is for bohmian mechanics. circumstantial evidence can be on a spectrum from weak to strong...

Semiclassical

I wouldn't even call it that. it's a plausibility argument

for me, I'm perfectly happy with Bohmian mechanics as an interpretation of QM (not an alternative theory) b/c I'm more interested in getting new predictions out of old theories

That Bohmian mechanics as such is experimentally equivalent to standard QM is to me a feature, not a bug

vzn

the hardcore physicists have said in here on record that new interpretations are hopeless/ near worthless for getting "new predictions". but all this sometimes reduces to semantic games similar to moving goalposts. if an interpretation advances far enough to come up with new predictions theyll say its not an interpretation any more.

Semiclassical

to be clear, when I say "new predictions out of old theories" I don't mean "getting different predictions out of old theories"

I mean applying the old theory in places where it hasn't been done before

(I say this somewhat optimistically, since I feel like nothing I ever did in grad school ever had much to do with predicting something in the real world...)

vzn

my judgement is that PWH is strong circumstantial evidence for bohmian mechanics + fluid dynamics in a sense. its too coincidentally aligned to be random or a mere meaningless curiosity. and yes it is inherently revolutionary even at this early stage because it definitively/ experimentally refutes the copenhagen interpretation...

Semiclassical

I think my quibble is with the word 'evidence' in there.

I'll assent that the existence of pilot wave hydrodynamics is an interesting fact and one which suggests further investigation.

vzn

18:42

← lol reminded of expr "damned by faint praise"

Semiclassical

The thing is that, experimentally speaking, pilot wave hydrodynamics is at least two steps removed from what can be experimentally observed

you start by going from standard QM to Bohmian mechanics, in which case one resuscitates the concept of a particle having a well-defined trajectory

bolbteppa

That last statement is to literally deny normal QM btw

Semiclassical

eh, not really. you get the same predictions, since you're only in a position to measure the particle once along a guided trajectory

once you've measured it, there's no reason to take it to be along that same trajectory

(and then life is complicated because the wavefunction isn't the same as it was before the measurement etc etc)

bolbteppa

Quantum mechanics begins from Heisenberg's uncertainty principle which says there is no concept of the path of a particle, if a well-defined trajectory did exist in any sense then classical mechanics would have to describe it, because all one needs is position and velocity and the force law in order to generate the paths, so if we take this seriously then Bohm is saying the classical force is simply being obscured, which I bet nobody actually claims

Semiclassical

well, given that the particle in Bohmian mechanics doesn't satisfy Newton's laws of motion

bolbteppa

18:49

Why shouldn't it satisfy Newton's laws if a path does exist

Semiclassical

Because it's not a classical particle subject to classical mechanics

That's simply not what the theory attempts to be.

bolbteppa

But what does that mean, if the path exists then we absolutely should be able to describe the path by some differential equation exactly

Semiclassical

Well, here's the usual sketch I give.

Suppose you have a (let's say 1D for simplicity) quantum system for which you can consistently prepare an initial state $\Psi(x,t=0)$.

bolbteppa

Where did the wave function come from

I must have looked at 50 different intro's to Bohm stuff and not one of them even attempts to do anything but axiomatically introduce this stuff out of nowhere

Even his paper assumed normal QM

Semiclassical

Sure. That's sorta the point, in my opinion: You start with stuff you already know about---i.e. being able to prepare an initial wavefunction---and go from there

bolbteppa

18:53

Ah man

Semiclassical

Why do you think I harp upon it being experimentally equivalent?

Balarka Sen

@Sid Belgian vs Japan is hilarious

bolbteppa

But that's not a theory, that's just saying 'look, this function spits out the results, I have no idea where it came from, in fact I stole it from a theory which makes sense and am pretending this is an alternative theory'

Balarka Sen

Best match 10/10

Semiclassical

But let me follow your point a certain ways. Suppose I claim that I'm able to prepare a quantum system to wavefunction $\Psi(x,t)$ for $t\geq 0$

i.e. I can prepare the initial state and I know the Hamiltonian therefore know the time evolution of the wavefunction

bolbteppa

18:55

Literally the only reason one can take quantum mechanics seriously is because it definitely claims the very variables of classical mechanics themselves cannot be specified - they can only be specified in some approximate sense which is why we experimentally measure paths in everyday life etc - but it absolutely destroys classical mechanics before it starts, if it didn't it couldn't be a theory or couldn't be different from classical mechanics

Semiclassical

How do I validate my claim?

bolbteppa

Bohm stuff seems to just introduce what one derives in normal QM from this thinking and then pretends it's a theory

Semiclassical

The simplest way I know how to validate the wavefunction at t=0 is to repeatedly prepare that system in the same way and then measure the particle location

bolbteppa

I don't appreciate the subtleties of Bohm stuff and would be happy to see it making sense, but it seems particularly egregious to claim a path actually exists while at the same time using the results coming from a theory whose tools derive explicitly from the assumption paths don't exist

Semiclassical

I'll agree that no classical notion of path is possible, i.e. that the particle is subject to classical mechanics. But again, that's not how it works.

But there's really no point in bothering with this conversation if you're going to dismiss the possibility of it being meaningful out of hand

So I can either continue with the story, or I can stop talking

bolbteppa

19:03

One set of notes I have defines a configuration $Q(t)$ and then actually defines it's derivative as $dQ/dt = \dots$, I mean that is already an unforgivable sin, even defining position and velocity is enough to say classical mechanics holds

Semiclassical

uh, that's crap

it's classical mechanics if $dq/dt=-\partial H/\partial p$ and $dp/dt=\partial H/\partial q$

bolbteppa

It would be great if it was meaningful, we'd all get to sneak in classical intuition into QM

Semiclassical

But Hamilton's equations don't hold for a particle along a Bohmian trajectory, so again: no, it is not a classical particle subject to classical mechanics

you can argue that the fact that said particle doesn't obey classical mechanics is a reason to dismiss it

heather

i thought i might be interested in taking the AMC 10 next year, so i found some practice problems to work through. Summary: i can't math. =P

methinks i have some studying to do.

Semiclassical

but then you're basically having it both ways. you're insisting that the only meaningful notion of particle trajectory is the classical one (e.g. hamilton's equations) while simultaneously saying that since QM != CM that particle trajectories are impossible

Anonymous

19:07

@heather What is AMC?

heather

@Blue american math competition. with AIME, it produces american olympiad team.

but it's also just a math competition.

vzn

pilot wave hydrodynamics has both newtonian and nonnewtonian elements. it can be seen as a hybrid. en.wikipedia.org/wiki/Non-Newtonian_fluid a key concept is emergence of higher level laws out of lower level ones.

Anonymous

@heather Oh, sounds like the national level counterpart of IMO

Anonymous

Cool

heather

@Blue nope, lower than national

Balarka Sen

19:10

@Sid 2-0!!!!!!!

heather

if you want to qualify for IMO as an american:

1) do well on the AMC, then you qualify for the AIME

2) do well on the AIME, then qualify for the american math olympiad

3) do well on the american math olympiad, qualify for IMO american team, go to summer camp, then go to IMO

Balarka Sen

IMO sucks IMO

heather

lol

Anonymous

@BalarkaSen This year's world cup is really weird

vzn

@bolbteppa lol sounds like QM/ copenhagen interpretation to me. :P

Semiclassical

19:11

To be clear, the reason I find bohmian stuff interesting is not because I necessarily find it's notion of trajectory to be compelling

Anonymous

All the unexpected things happening at once :P

heather

anyway, example AMC (AMC 10 for 10th grade and under, 12 for 12th and under, but there's one AIME for everyone). much more problem focused as opposed to proofs.

Semiclassical

i find it interesting because it's an internally consistent trajectory-based story (and the simplest one one can extract from QM)

Balarka Sen

@heather You asked an interesting problem about inverting gradient over a discrete data a few weeks ago. Did you see my proposal?

I thought your question was really nice.

Semiclassical

If I know the wavefunction in some spacetime region, then I can write down the probability density $\rho$ and the probability current $\vec{j}$

heather

19:13

@BalarkaSen let me check

Semiclassical

and then the ratio $\vec{j}/\rho$ defines a velocity field throughout that spacetime region

heather

oh in chat

Semiclassical

The moment you say that there's a wavefunction appropriate for that spacetime region, then those quantities exist and therefore that velocity field exists

and that velocity field will have streamlines.

Balarka Sen

Yeah.

heather

here:

in Mathematics, Jun 19 at 15:58, by Balarka Sen

So if you know the value of your inverse-gradient at some point $x \in \mathscr{D}$ of your data set, say the value is $f(x) = p$, for any other point $y \in \mathscr{D}$, $"\int_x^y \nabla f = f(y) - p"$, from which you recover $f(y)$.

?

Semiclassical

19:14

as such, any wavefunction has a set of trajectories. now, you can argue that those trajectories cannot and should be interpreted as real ones

but mathematically those trajectories are there the moment you suppose that there's a wavefunction on that spacetime region

Balarka Sen

Yep @heather

heather

i thought it was an interesting way to go about it, but i don't know if i know the inverse gradient at some point...i was trying to solve a problem for my dad, and an inverse gradient was required.

Semiclassical

They're a mathematical part of the theory. As such, they're interesting to me.

enumaris

welp, I finished writing my updates to the code for today...can't test it until the current analysis finishes...no idea how long that will take...

Novalium Company

youtube.com/watch?v=JGXdi7XcQi8 At 17:36. What I don't understand is why if we want to keep the voltage at 3V across R2 (and Rload), why the resistance of R2 has to be 100 times less than the resistance of the load?

bolbteppa

19:18

Classical mechanics is all about specifying the position and velocity of a particle at a given instant, and then using the laws of motion to predict where it will be at all future instants - this is done by solving second order differential equations which require the specification of an initial position and an initial velocity.

Semiclassical

@bolbteppa Yes.

But, funny thing, the momentum of a particle in the Bohmian picture isn't subject to a second-order law.

it's subject to a first-order law.

bolbteppa

If a path, any path, exists, we should, somehow, be able to specify the initial position and initial velocity, and then describe that path by some differential equations with those initial conditions, clearly not the simple $F = ma$ - so if you tell me Bohm is saying a particle does not satisfy 'classical' equations of motion, that's fine, but you're just telling me we don't know the laws governing why the particle follows the actual path it takes.

In other words, you've just said the current force laws governing the motion are wrong and we need to modify the values we give to our force functions. This is all just mathematics, this is simply undeniable... The radical thing QM does is it says we simply can't define the variables in the first place, which is why it makes so much sense, otherwise Newtonian mechanics would be fine.

Semiclassical

Fun point: In the Bohmian story, you don't get to treat the initial position of a particle and its initial velocity as independent quantities.

Balarka Sen

@heather Gotcha. Very interesting problem nonetheless.

vzn

@bolbteppa in PWH the particle follows a path that is an interaction between its surface and the wavefn. the interactions are (maybe) locally newtonian but globally follow "emergent" and hence nonnewtonian trajectories. it is incorrect to say "we dont know the laws governing why the particle follows the actual path it takes."

Semiclassical

19:22

If you know the wavefunction at t=0 and the initial position of the particle, then the velocity at t=0 is determined.

The very notion of being able to specify the initial position and the initial velocity isn't an option in the Bohmian story.

bolbteppa

I'll try to say it again: If particles follow paths, then all we need to do is solve some differential equations to find those paths, if we measured the initial position and velocity for a given particle, we'd then be able to say that $F = ma$ is simply wrong, or rather, the value of $F$ we would expect from normal physics is giving us wrong answers, we need to modify how we specify $F$.

There is literally zero need for a wave function theoretically. Now, we can randomly introduce wave functions as a way of saying 'we don't know the true $F$, so lets use a probability function to guess where it should be'. That's all fine, but it's literally just classical mechanics but saying we are measuring the wrong $F$'s

Semiclassical

You keep insisting that the story being told is the one y ou're telling

bolbteppa

I'm happy to be corrected man

vzn

@bolbteppa try looking into PWH. its purely classical (locally) in a sense. suggestion, try to understand it 1st without thinking of its QM implications at all.

bolbteppa

Why can't you specify the intial position and velocity in Bohmian mechanics?

Semiclassical

19:26

PWH != Bohmian mechanics is something I"ll insist on here

with the relation between them being that Bohmian trajectories arise as averages over subquantum fluctuations of particle motions in PWH

because the moment you specify the initial position and the evolution of the wavefunction, the velocity is fixed

enumaris

what about that paper saying there are no particles tho

only fields

Semiclassical

as an example, for any stationary bound state, the probability density is constant in time and therefore the probability current is everywhere zero

enumaris

arxiv.org/pdf/1204.4616.pdf

vzn

Thailand cave rescue: Boys found alive after nine days / bbcnews

Semiclassical

as such, as a matter of principle, the only allowed velocity of a Bohmian trajectory for such a stationary bound state is zero: $dQ/dt = j/\rho=0/\rho=0$

enumaris

19:29

@vzn any news on how hard/easy it will be to get them out?

danielunderwood

@enumaris Isn't that one of the key thoughts of QFT?

(I know they aren't speaking of QFT here, but still)

enumaris

Generally QFT still deals with particles and fields with particles being field excitations

bolbteppa

But "it also postulates an actual configuration that exists even when unobserved. The evolution over time of the configuration (that is, the positions of all particles or the configuration of all fields) is defined by the wave function by a guiding equation." right?

enumaris

I have yet to determine if the paper is just reiterating that in a more forceful way

Semiclassical

Sure. But you have to know what the wavefunction is in order to say what the velocity will be

bolbteppa

19:30

It even says the configuration actually has a velocity, $dQ/dt = (1/m) \mathrm{Re}\dots$

enumaris

I know there's some issues w.r.t. particles in QFT in curved spacetimes

bolbteppa

Ok

enumaris

in the sense that in curved spacetimes different observers see different fields, so some observers would see particles and some wouldn't (e.g. Unruh effect)

bolbteppa

So we can't find the velocity for some reason, but it's still there

Semiclassical

The most you can ever do in Bohmian mechanic is infer a velocity

vzn

19:31

@enumaris they were found by divers, it looks like the cave flooded substantially (as it has been known to do this time of year) after they entered. the extraction looks like it could be very difficult because maybe they can only be remotely reached by scuba divers right now. found alive is the 1st miracle, all extracted safely would be the 2nd. now playing out.

enumaris

D:

I hope they can get them out alright...

Send them my energy つ ◕_◕ ༽つ

vzn

the story reminds me a lot of trapped chilean miners a few years ago. guess 2010. another riveting intl viral rescue story that played out online.

enumaris

being trapped in a cave must seriously suck

oh wow, belgium has tied it up lol

Semiclassical

better than being dead in a cave tho

vzn

← currently trapped in a deep dark theory cave, no sign of forthcoming rescue

Semiclassical

19:35

which i'm sure is what they were worried about

enumaris

@vzn I'll rescue you, just wait until I've learned all of physics and have come up with a TOE

For sure it's good news that these kids have been found

vzn

@enumaris lol thx man youre awesome nicest thing anyones said to me in weeks in here :P you could rescue me right now if you want to read my blog :P

enumaris

U gotta wait for me to learn the rest of physics first

vzn

@enumaris oh geez its cute but maybe stop pretending you didnt get a physics phd :P

Semiclassical

Hrm

I coulda sworn there was a non-paywalled version of this article: advances.sciencemag.org/content/2/2/e1501466.full

enumaris

19:37

Link me to one blog post, I'll read one

can only do one at the moment

vzn

@enumaris lol hope your firewall doesnt nix it, would be surprised if it didnt :P vzn1.wordpress.com/2018/05/25/fluid-paradigm-shift-2018

@enumaris 2nd that! always nice to see signs of actual camraderie/ empathy/ humanity etc in here :P

Semiclassical

the upshot being that, in order to deduce the bohmian velocities, what they have to do is many many weak measurements of momentum

enumaris

it's not blocked

vzn

@enumaris thx god for small miracles =D (my own co blocked it for a long span once!)

Semiclassical

(i.e. do a weak measurement of momentum, then a strong measurement of position, ad infinitum)

That should already signal a big difference with classical mechanics. if i have two trajectories which happen to have the same position, there's no reason to expect that they'll have the same velocity

vzn

19:42

@enumaris nice find thx for sharing quite notable adding it to my ~~criticized~~ huge list :)

Semiclassical

hence this measurement technique makes little sense classically. it only has a hope of making sense if the Bohmian velocity is dependent upon the Bohmian position

to put it a little differently: in classical mechanics, the most proper notion of path is not 'path in configuration space' that of a path in phase space

vzn

> There are no particles, there are only fields
> Quantum foundations are still unsettled, with mixed effects on science and society. By now it should be possible to obtain consensus on at least one issue: Are the fundamental constituents fields or particles? As this paper shows, experiment and theory imply unbounded fields, not bounded particles, are fundamental. This is especially clear for relativistic systems, implying it's also true of non-relativistic systems. Particles are epiphenomena arising from fields.

Semiclassical

classical particle trajectories can and generically do cross in phase space, since you can have two particles at the same position with different velocities

but that is emphatically not true for Bohmian trajectories. the bohmian trajectories deduced from a given wavefunction are flowlines

as such, they don't cross in configuration space

(That property alone---Bohmian trajectories don't cross in configuration space---should really be a red flag all by itself that it's not a Newtonian theory, since classical paths can and do cross in configuration space)

enumaris

halfway through lol

Semiclassical

i mean, the law for the velocities is dQ/dt = j/rho. but j and rho are determined by the wavefunction, and are not affected at all by Q. the exception here is if you actually measure Q: the wavefunction after measurement certainly depends on the value of Q observed. but that's just regular QM

Balarka Sen

19:52

@Sid Woo, Belgium recovers

This was actually a great game

bolbteppa

That's fine, you can only infer a velocity, we know it's there but we can't find it - but if a path, any path, actually exists, there exists some differential equation which spits this path out, just by the theory of differential equations. Newton's classical mechanics claim's to give the rule by which you find this true path,

Semiclassical

you keep saying that.

bolbteppa

i.e. the rules for setting up the ode's, you determine the forces then just use these values for the force in some differential equations, solve them, and then if you've measured the initial conditions, voila, you know how everything works from then on.

Literally the only thing that we can change in this picture is the value of the force function, since now the 'force' is just a portion of a second order ode which spits out the actual path (if we specify the true initial conditions which theoretically exist even if we can't find them for some reason).

Thus, if we allow even the idea of a path, all we can say is either that Newton or normal intuition is spitting out the wrong value of $F$ because we're not measuring what we should, but that maybe we could later find the 'true' way to specify the real F, or that the truth is there are no rules describing how to specify the $F$, it's arbitrary, different in every situation, and we can only specify it on average (quasiclassical limit).

But still, theoretically we should always be able to think up some 'generalized force' which accounts for the discrepancy for each case and set up a differential equation representing this path, not really a theory to have billions of different ode's but that's the logic of this.

Thus we can simply replace all this with probabilities only because experimentally we know classical physics fails and for absolutely no reason we set up an equation which magically works and spits out the right answer, Schrodinger's equation. What am I missing?

vzn

@bolbteppa the sch. eqn has deep ties to classical (wave) mechanics + fluid dynamics.

bolbteppa

@vzn what does "classical (wave) mechanics" mean, wave mechanics is quantum mechanics

vzn

19:55

@bolbteppa try looking into so called derivations of sch. eqn.

bolbteppa

@vzn I have

15

Q: Variational Derivation of Schrodinger Equation

In reading Weinstock's Calculus of Variations, on pages 261 - 262 he explains how Schrodinger apparently first derived the Schrodinger equation from variational principles. Unfortunately I don't think page 262 is showing so I'll explain the gist of it: "In his initial paper" he considers the re...

quantum-mechanics wavefunction schroedinger-equation variational-principle complex-numbers

I posted a bunch of questions on this ages ago

Note I am quoting Schrodinger's original paper

Semiclassical

I'm doing the following off the top of my head, so if this goes wrong I won't be surprised

bolbteppa

Sure

Semiclassical

but, suppose you've got a classical particle subject to a damping force and a potential

enumaris

@JohnRennie welp, you were wrong, Japan is out :P

bolbteppa

19:58

15

Q: What is Quantization?

In classical mechanics you construct an action (involving a Lagrangian in arbitrary generalized coordinates, a Hamiltonian in canonical coordinates [to make your EOM more "convenient & symmetric"]), then extremizing it gives the equations of motion. Alternatively one can find a first order PDE fo...

quantum-mechanics lagrangian-formalism hamiltonian-formalism quantization complex-numbers

Semiclassical

So the equation of motion will be $m\ddot{x}=-b\dot{x}-\frac{dU}{dx}$

bolbteppa

@vzn read those

vzn

@bolbteppa this is not my exact area of specialty but try understanding how much of sch.eqn is similar to the classical wave eqn... you seem close. there is a nice "derivation" in baggott "meaning of quantum theory". there are also some other refs that tie it in to fluid dynamics, would take me awhile to dig all those up, but have cited at least 1 before...

Semiclassical

now, there are various limiting cases of that setup

in particular, you can consider the case of strong damping

enumaris

@vzn btw GUT is not a TOE

bolbteppa

19:59

@vzn here's an entire book trying to make sense of Schrodinger's method:

amazon.com/Probability-Schrodingers-Mechanics-David-Cook/dp/…

Semiclassical

in that case, one neglects the inertial term to get $b\dot{x} = -dU/dx$

vzn

@enumaris agreed, theres an interplay. but (as asserted in that blog) some of the distinctions are artificial.

Transcript for