The h Bar

Cows

7:13 PM

HAVE PAID OFF MY ASU DEBT::: CAN NOW RETURN TO SCHOOL>>> SENDING TRANSCRIPTS TO NEW COLLEGE AND . . . . MORE HEHEHE

ORDERING MY TRANSCRIPTS NOW

YAAAAAAAAAAAAAAAASSSSSSSSSSSSSS!!!

Anonymous

@Cows Congrats!

Cows

@Blue Thank you

Anonymous

What are you applying for?

Cows

well, I'd be doing physics hehe

I have two classes to take then, I will be applying to grad school too hehe

Anonymous

@Cows A masters degree I suppose?

Anonymous

7:15 PM

Or a PhD?

Anonymous

@Cows Sounds great :)

Anonymous

Good luck

Cows

It depends, I used to want to jump to a phd directly, but now some knowledge about life I am not opposed to a masters first hehe

Anonymous

I can understand :P

Cows

I am also going to try to get a position at an aerospace company I've been wanting, but could not because I needed to show my transcripts

I just called the manager (who happens to be my friend ) and they are waiting heheh

Anonymous

7:17 PM

Umm, won't aerospace company+grad school get a bit too hectic ? :P

Cows

not if you figure out a way to make your job be your school research project ;)

@Cows nice

@bolbteppa thank you

@Cows congrats =D

@heather thank you

Anonymous

7:23 PM

Lol, apparently I found a bug in the SE bounty system. Deleting the question on which you gave a bounty, and subsequently triggering a profile deletion, and then again canceling the profile deletion gives you back your lost rep points (experimental conditions - the bounty was given using my account's associated rep on Math Overflow; the account had only 1 question).

Anonymous

(which basically implies I can give out an unlimited number of bounties on any SE site, as long as the profile has no other associated questions)

ACuriousMind

7:42 PM

@Blue You should report that to SE as a bug and probably not abuse it.

enumaris

So I got back into MTG over the weekend...it's still quite fun, even though I'm not familiar with any of the new mechanics

Anonymous

@ACuriousMind If you're both an AI and a saint, it will be difficult to conquer the world ;) (fwiw, yes, maybe I'll write about it on Meta SE)

enumaris

He is no saint, he just wants bugs reported so he has fewer variables to worry about when he eventually takes over

fewer variables = easier to calculate to the endgame

Mithrandir24601

@Blue I do know someone doing their PhD at NASA :P

Anonymous

@Mithrandir24601 NASA gives out PhDs? :P

Anonymous

7:49 PM

Or maybe just a project ?

Anonymous

BTW I thought @Cows was more into mathematical physics (?)

Mithrandir24601

@Blue It's in collaboration with one of the London Unis, actually working on one of D-Wave's machines, just to make it sound more impressive :P

Anonymous

@Mithrandir24601 Strange....NASA....London uni......DWave...Does NASA have DWave machine ?

Mithrandir24601

@Blue yep!

Anonymous

Ah, makes sense

Mithrandir24601

7:53 PM

ti.arc.nasa.gov/tech/dash/groups/physics/quail

Anonymous

Cool. Tell your acquaintance to leak the DWave architecture details :P We already have so many sceptical questions about it on the main site

Mithrandir24601

@Blue We had to sign NDAs when we visited, so people wouldn't be impressed :P

Semiclassical

8:11 PM

One thing I really don't know about QC: how much of the work to be done is hardware vs. software?

heather

all of the above

=P

no, i guess i would say that software isn't really an issue yet because hardware isn't far enough along.

but @Mithrandir24601 could probably give a better answer than me =)

Semiclassical

Well, on the software side I include stuff like error-correcting codes

Mithrandir24601

@Semiclassical Yes, all of the above :P

Semiclassical

lol

Slereah

turns out Schwarzschild with a Gaussian gravitational potential isn't eminently solvable

Mithrandir24601

8:15 PM

@heather More or less, to be fair - the software won't really matter until/unless we get to the point of proving some kind of advantage (although 'software' will be required to actually do that)

@Semiclassical Um, yes? :P

Semiclassical

lol

Mithrandir24601

oh no wait, I've got it - true

Semiclassical

that bit always makes me lol

"No there is, or no there isn't?!?" "Yes."

Mithrandir24601

8:18 PM

true is right. Win for CS :)

Semiclassical

hah

heather

@Mithrandir24601 i thought that recent paper finally proved that

Semiclassical

The problem I have on the outside looking in re: QC is that the issues seem to either be about hardware, in which case I'm irrelevant since my knowledge of experiment is weak, or about software, in which case I'm also irrelevant since I know f*** all about computing.

heather

this thing

Mithrandir24601

@heather It's complicated - it's constantly changing (very slowly), so a new algorithm could come along and shift that either way a few qubits here or there

Semiclassical

8:21 PM

About the most someone like me can do is help educate what QM is about.

Anonymous

@Semiclassical But algorithms don't really fall in any of those categories. Also, a physicist can obviously work on theoretical models of quantum computing (it doesn't necessarily need to be experimental). Ideas come from everywhere

Mithrandir24601

@heather And if you write a quantum circuit for it, you can still solve it classically, just not efficiently

ACuriousMind

@Blue ...how do algorithms not fall into the category of "computing" or "software"?!

Anonymous

@ACuriousMind I was referring to "software"

Anonymous

If you stretch the definition of "software" a bit, then maybe

Anonymous

8:26 PM

I'm not completely sure in what context SemiC is using that

Semiclassical

A vague and ignorant one. :P

enumaris

meta question

ACuriousMind

I mean, an abstract algorithm is not a specific "piece of software", but designing efficient and correct algorithms is a classic part of what people who engineer software do.

enumaris

if you see a multi-part question which you have the answer for only a subset of the parts, are you supposed to just let it go and wait for someone to answer all the parts of the question

or do you answer the parts you know

try to fit partial answer into a comment?

ACuriousMind

@enumaris No, you are supposed to VTC the question as "too broad". If question asks subquestions that can be answered independently, it should be split into several actual questions.

Anonymous

8:27 PM

@ACuriousMind I thought you call people who invent new algorithms, computer scientists (?) :P Btw, that distinction is fading away nowadays especially with software companies doing research work

ACuriousMind

Additionally, partial answers are answers - you are perfectly allowed to put a partial answer as an answer.

enumaris

hmmm, but what if every part is actually quite inter-related.

ACuriousMind

@Blue That computer scientists design algorithms does not mean that software engineers don't!

enumaris

hmmm

Semiclassical

To be clear, I was including 'algorithms' and stuff like error-correcting codes under the umbrella of 'software.'

Hardware vs. software is a very broad distinction, though.

Anonymous

8:31 PM

@Semiclassical I think physicists can very easily pick up error correcting codes and all that stuff (in about a year)

Semiclassical

possibly, yeah

The stuff I've personally been curious about lately is more in the realm of Bell stuff

Anonymous

@Semiclassical That's more of quantum information theory :)

Semiclassical

Though I think it could be classed under the rubric of 'semidefinite programming'

I think I do have at least some understanding of how bipartite correlations work now, at least

(I'm sorta tempted to play around with tripartite correlations, if only b/c that's where GHZ comes in)

Anonymous

Sounds interesting. I yet have to pick up those parts. Information theory is indeed where the more juicy part lies (although not very relevant to attempts at practical implementations of QCs)!

Semiclassical

Right.

Naveen Balaji

8:43 PM

Hey guys! Please do check this out if free.

3

Q: Understanding the Fujitani-Ikeda-Matsumoto embedding

The minimal dimension $N$ of the ﬂat space in which the Schwarzschild metric can be embedded is equal to six, and the Fujitani-Ikeda-Matsumoto embedding is a method for embedding the Schwarzschild metric in a six-dimensional spacetime. There are two possibilities of the signature choices: $(++---...

Slereah

Solving the geodesic equation is always p. bad because it's basically always nonlinear

Anonymous

@Semiclassical Btw, if you get time, considering answering this (?) I'm not very satisfied with the existing answers

Anonymous

DaftWullie says square root of NOT gates are not allowed by classical probability

Semiclassical

hmm

Anonymous

Brian R. La Cour says that the main difference is that for classical coins given an initial state (precisely) we can exactly determine the outcome. But that's not possible for a qubit

Semiclassical

8:52 PM

what I'm curious from there: they give the example of using a half-silvered mirror

which works fine for a photon

but the usual example I use for a qubit is a single electron going through a Stern-Gerlach device

and I'm not sure what the analogue would be

I think this reveals some ignorance on my part, though: I don't know off the top of my head how to get non-classical behavior out of a single qubit. I know how to do it with two qubits, e.g. doing spatially separated measurements on a singlet state.

Anonymous

@Semiclassical Exactly my point. I'm not sure how QM behaviour can be distinguished from CM behaviour given just a single qubit

Semiclassical

Same.

I imagine there is, but I don't know how it works.

Anonymous

35

A: What makes a theory "Quantum"?

I think this is a subtle question and I think it depends somewhat on how you choose to represent quantum mechanics. To see one extreme of this, consider the viewpoint put forth by Kibble in [1]. For simplicity I will be thinking of finite-dimensional quantum systems here; there are some subtletie...

Semiclassical

I do think I've got a pretty nice answer for the two-qubit case, though, one which ties back to Bell's original inequality.

Anonymous

This answer says something similar (although I haven't been able to comprehend it completely)

Semiclassical

8:58 PM

Which I'd sorta like to test out [ulterior motive drop :) ]

Anonymous

"

All of this is to say that it seems difficult (and maybe impossible) to try to find a single distinguishing feature between classical and quantum mechanics without considering composite systems, so if that is what you want, I'm not sure I have an answer. If you do allow for composite systems though, it is a pretty unambiguous distinction. Given this, it is perhaps not surprising that all the experimental tests we have which demonstrate that the world is quantum and not classical are based on entanglement."

Semiclassical

Yeah.

Anonymous

@Semiclassical Ah, I think it would interesting to have that on our QCSE site. You can consider writing a self-answered post regarding the two-qubit case. I'll be interested

Semiclassical

I'm not going to claim there isn't an experiment that works for a single qubit, but I don't know what it is.

Anonymous

Same here

Semiclassical

9:03 PM

Hmm. I guess the analogue of the half-silvered mirror should be the application of a magnetic field

i.e. precession

Mithrandir24601

@Semiclassical A single photon source would do the job - show that it has sub-poissonian statistics

Semiclassical

Sure, but I want something with electron spins since that's my bread and butter

grand_unifier

I am currently on a 4-year Mphys undergraduate degree (finished my 3rd year) in the UK. Although I am not sure yet, I am very interested in fundamental particle physics (High Energy Physics). During the summer I am determined to become familiar with the concepts of QFT in order to make sure I like the subject. My question is, how likely is it that after a Ph.D. in Theoretical Physics, one can work in a university(I figured very low?)?

What is the path to becoming a lecturer?

Semiclassical

I think the magnetic field example is the right one, since an electron spin in a constant magnetic field will precess around the field axis if I'm remembering intro QM right

so there should be a direct analogue using electron spins alone.

Mithrandir24601

@grand_unifier It depends on many, many factors, the biggest one is probably the PhD and the second biggest is probably luck. On average, on what I do (quantum engineering, supposedly), about 50% have the opportunity to do at least a short PostDoc, while maybe 10% would actually do a full 2-3 year PostDoc

grand_unifier

9:09 PM

@Mithrandir24601 I realize it's not a fully answerable question, I just want to hear as many opinions as possible from people who know at least smth more than me.

Semiclassical

for comparison, the matrix $U$ given by DW is just $U=\sqrt{2} e^{i \sigma_x \pi/4}$

Mithrandir24601

@grand_unifier To be honest, I wouldn't worry about it - do you want to do a PhD over whatever the alternatives are? If you don't want to, it could be a terrible decision to do one. The reason I say that ~50% get the opportunity is because a fair number don't take the opportunity and get an industry job or something instead

Semiclassical

corresponding to the application of a magnetic field in the x-direction for an appropriate period of time

(a quarter period I guess)

Mithrandir24601

@Semiclassical But can I make something classical that gives the same statistics?

Semiclassical

There be the question.

I want to say that the point goes like this: If you measure an electron as spin up on the z-axis and then apply a magnetic field in the x-direction for a quarter period, then you're as likely to have Sz=up as Sz=down

Mithrandir24601

9:14 PM

I think this is where the real issue lies - you can compare something quantum to a specific classical example and say 'oh you can't do this with this specific example' but as QM is all about the statistics (you've only got one shot at measuring, after all!), it's doesn't matter if classical systems behave differently, only if you can get them to have the same statistics or not

Semiclassical

So that precession apparently eliminates all your information about up and down.

But if you were to have done two such precessions to begin with, then that amounts to flipping the spin

Mithrandir24601

Hence, it always boils down to the same thing: Bell/CHSH inequality/correlations/negative quasi-probabilities/sub-poissonian statistics etc. which are all different aspects of the same fundamental principle

Semiclassical

in which case you can be sure that you'll measure spin down

So you have an operation which, when done once, eliminates all knowledge of the original spin

But if done twice (prior to measurement) it's guaranteed to reverse the original measurement

But, eh, that doesn't really seem so different from a coin flip. If you could track it in mid air, and always measure it when it's in the same orientation it started with, you'd be guaranteed to get the same measurement you started with.

So there doesn't seem much non-classical about that.

By extension, I find myself less impressed with the example of a M-Z interferometer than I initially was.

Mithrandir24601

@Semiclassical It's good for a HOM-dip though

Semiclassical

(I have a little more in mind with the 'tracking a coin in midair' example than I've said, i'm just being lazy and not saying so)

Mithrandir24601

9:24 PM

(of course, HOM also requires 2 photons...)

Semiclassical

yeah...

The other part of this is that the point is not so much "can you do it without QM" but "can you do it under certain constraints"

The nice thing about the Bell inequality setup being that the separability constraints seem so obvious

bolbteppa

@grand_unifier competition for a phd is insane

Semiclassical

I suspect the spin equivalent to HOM would be: entangle two electrons in a singlet state and separate them. for spin 1, measure its z-component immediately. for spin 2, apply a magnetic field for a time T and then measure the z-component.

do that a bunch of time to get coincidence statistics

clarification: when T=0, the spins will always be different. there'll be time T=T' such that the state returns to the initial state up to a minus sign. at the half-time T=T'/2, the spins will always be the same. at the quarter time T=T'/4, then the spins will be uncorrelated.

Mithrandir24601

9:54 PM

@Semiclassical Allowing for changes in time would make it more complicated though - it would have to be performed with both spins very far apart

Semiclassical

Yes, well, no one said I wanted this to be practical :P

grand_unifier

10:11 PM

Just a follow up on my previous question. What would you guys do to maximize your success rate in becoming a theoretical particle physics lecturer? From maximizing your chances to get a PhD position -> Post-doc position -> Research Fellow -> Lecturer? -> (I wish I ever get here).

Physics Meta

10:50 PM

-1

Q: Allow Homework Help Questions

math.stackexchange welcomes homework help questions with open arms, but physics.stackexchange doesn't. Every physics homework help question is immediately put on hold as off-topic even if OP provide significant work, but reached the wrong answer. For example: Homework: Potential Energy I thi...

discussion feature-request homework-and-exercises policy

Physics Meta

11:13 PM

0

Q: Harmful edits: which action?

If I find an edit which caused more harm than help by a new user, and which was subsequently approved, is there any action I should take, in addition to fixing the edit? (And in this case, should I just fix the edit, or first rollback and then fix?)

support editing suggested-edits

Transcript for