The h Bar

Tobias Fünke

07:23

they just proceed. and the more answers I read, the more I am convinced that it is AI...

mostly in their recent answer about Haag's theorem. They repeat the same stuff like three times

Ryder Rude

hi

naturallyInconsistent

08:00

@Loong it's half in jest

Ryder Rude

08:31

@GroveRover Prof Legolasav on PSE knows LQG. u can ask him

Rovelli has some lectures here youtube.com/…

GroveRover

@RyderRude I forgot that a professor I know well worked in Rovelli's group so I'll ask him

Ryder Rude

@GroveRover oh

GroveRover

@RyderRude Thank you but I can't follow lectures. Need to fix that before the school lol

Ryder Rude

there r two brands of LQG. i think they're called covariant and canonical LQG. ur camp is about the former

in canonical LQG, they use some techniques like constructing gauge invariant Hilbert spaces. this technique is common in gauge theories. Rovelli uses this in the lecture

string theory also uses this technique in constructing the Hilbert space of the closed string

but Rovelli also considers a triangulated version of spacetime, which might not be useful in other quantum gravity approaches

like approximating space using tetrahedrons. i think it is specific to LQG

GroveRover

@RyderRude It's like Einstein and ADM formalism?

(Quantum version clearly)

Ryder Rude

08:38

@GroveRover the latter is based on canonical quantisation. idk about the former. but the former is the one that has been shown to reproduce GR

GroveRover

@RyderRude nice thanks

Ryder Rude

@GroveRover so yes, the latter is like the ADM formalism. i think they first make a change of variables from ADM to Ashtekar variables and then do some ad-hoc stuff and then quantise it

ad-hoc stuff like complexifying the metric at the classical level

and they treat time as a continuous parameter, but derive quantisation of space. so it seems ugly

but the covariant approach may be less ugly

Tobias Fünke

09:28

Have you ever checked your Erdős number? What would be an equivalent in physics/ your field of research?

antimony

10:51

wonderful thanks @TobiasFünke really appreciate it. you're definitely the best analyst-therapist around ;)

ACuriousMind

@TobiasFünke The problem with that answer wasn't so much the repetition as that it's mostly nonsense. Now if only all the LLM spammers did me the favor of generating answers on topics I understand well-enough to tell they're gibberish... :P

Tobias Fünke

@ACuriousMind yes

glad that you took care of it--and also of the question ;)

@antimony ;)

oh. the other user I've meant is also "gone"

oh dear

jrobins

11:10

hello friends, I recently posted a question here and i'm super happy with the responses. Yesterday, I got a well written elaborate answer on my first question. I wanted to read it with a fresh mind after sleep. However, now the answer is gone!!! I'm so upset!!

Is there any way I can recover this answer?

ACuriousMind

@jrobins The answer was deleted because it was very likely to be AI-generated.

Tobias Fünke

the user was banned for plagiarism (most probably copy-paste AI generated content). So if you want to get a similar answer, I'd suggest to put your question to chat gpt lol

jrobins

wow, thank you so much for the response. Would be bad if comments are AI generated. I'm hesitant that the way I use chatgpt would generate the same answer.

I understand higher reputation accounts can see deleted comments? Is there a way I could possibly see the answer myself again? Maybe someone can send it via a private message? I would be super grateful

something as easy as a screenshot would suffice, copy pasting tex could be problematic sometimes

Sine of the Time

11:32

@TobiasFünke mathscinet.ams.org/mathscinet/freetools/collab-dist in case you want to check your Erdos number

ACuriousMind

@jrobins You need 10k reputation to see deleted posts. I'm hesitant to provide the content of an answer that was deleted because the user violated site policies, and for this specific answer I cannot really judge its accuracy.

Tobias Fünke

@SineoftheTime yeah I know. I just did this morning, this is why I asked hehe

jrobins

@ACuriousMind I understand. I actually found the answer in an old tab, I will read it now. Is there a way I can help the authors case if I judge the answer is not AI generated or plagiarised? Given what I read I find it highly unlikely: I tried using AI for this but it's just not smart enough to give such an answer.

Tobias Fünke

my Erdős number is 5, my Feynman and my Heisenberg numbers are 6, my Mermin number 5 :d funny

ACuriousMind

@jrobins I agree it's probably not the result of just putting your question into standard ChatGPT, but looking at the overall pattern of answers (e.g. amount of text written in very short time frames, wildly varying topics, certain patterns of repeating phrases from the question or previous paragraphs verbatim) I'm satisfied these answers are not (all) handwritten.

Tobias Fünke

11:47

isn't there a version of chat gpt which is quite good in natural sciences and math?

but you have to pay for it, if I remember well

qwerty

11:58

I'm disappointed that deleted does not truly mean deleted. I understand for moderators maybe but not high rep users.

ACuriousMind

@qwerty deleted posts have to be visible to high-rep users, since high-rep users can vote to delete and undelete posts

Tobias Fünke

does "undelete" happen actually?

PM 2Ring

Members with 10k+ see the deleted answers on any question page. But we can't search for deleted posts (apart from our own), so we can't easily see deleted questions, unless we have them bookmarked.

ACuriousMind

@TobiasFünke very rarely

PM 2Ring

Well, stuff is usually deleted for a pretty good reason. And even if the author edits the deleted post, the workflow around undeletion is a bit clunky, so undeletion is fairly rare.

Also, regular users can't vote to undelete stuff that's been deleted by a diamond mod.

@qwerty In extreme circumstances, post revisions can be redacted, so nobody can see them. But that's normally only done when the post contains info that shouldn't be public, like real passwords or API keys.

Of course, for that to be effective, the info has to be deleted swiftly. There are a lot of scrapers harvesting stuff off Stack Exchange.

@ThomasOwens I'm an SRE so I focus on keeping the app running optimally, feature requests would go to a dev team. I'm giving rationale from my team's point of view. As far as hard numbers, in the last hour there were around 3.88mm requests to Stack Overflow. Around 984k requests were from non-verified bots. Without digging too deep, I'd say the majority of those requests are from scrapers and undesired. As far as respecting robots.txt, it'll be abundantly clear which bots are respecting it and what's not once we make the change, I can't give more details beyond that. — Josh Zhang ♦ Dec 3 at 21:11

@ACuriousMind I would really appreciate it if you took a look at this suspected case of LLM plagiarism, too...

13 hours ago, by PM 2Ring

This one? https://physics.stackexchange.com/users/315779/badri-narayanan Totally quiet for 3 years, then 4 long answers in 18 minutes!

jrobins

12:36

@ACuriousMind I agree the answers are AI-like and probably not entirely handwritten. It must be some good AI model usage since the answer mostly makes sense. Actually, I think I learned more than the current best voted answer to my question. I think it's upsetting it's not visible to me. Perhaps a flag 'probably used AI' would've been better than deleting the post.

Herr Feinmann

13:38

@qwerty yeah, it hurts :P

PM 2Ring

13:56

in Python on Stack Overflow Chat, Jul 4, 2016 at 11:08, by PM 2Ring

The good thing about 10k rep is that you can see all the deleted answers. The bad thing about 10k rep is that you can see all the deleted answers.

Tobias Fünke

xD

naturallyInconsistent

lol

PM 2Ring

It's not so bad on Physics.SE. On popular questions, there are often a bunch of similar bad wrong answers, and occasional troll posts. On SO, on big questions there can be dozens of deleted rubbish posts.

Some of those may be technically correct, but they're basically just dupes of older answers, posted in an attempt to earn some rep. We don't mind the occasional dupe, especially if it's unintentional. But having a dozen near-identical answers just adds clutter.

Silly Goose

14:19

@TobiasFünke i think mine is undefined :P. If you count undergrad advisor my preskill number is 3 and my weinberg number is 4

Herr Feinmann

@PM2Ring lol

It also depends on the reason why you delete it. For example for me having something wrong on my profile feels like "tarnishing my criminal record"

Making what I say appear less reliable at least as a first impression

PM 2Ring

@HerrFeinmann Yeah, I understand that. But it's much better if you can actually edit the answer to make it correct. But sometimes that ends up turning your answer into a clone of an existing answer.

Herr Feinmann

What about questions? :P

My way to go is usually editing it to make it right and then delete it. Just like in real life I sometimes correct everything before trashing some paper :P

PM 2Ring

14:35

Also, new members often delete their downvoted questions or answers, to improve their "record". But the Stack Exchange ban algorithm doesn't see it that way: deleting a bad post doesn't make the badness go away, and deleted bad posts contribute to a ban. This can come as quite a shock...

Once you have a reasonable track record, it's very unlikely that you'll get a question or answer ban. But it does affect a lot of newbies, especially on SO, and some of us feel that the system documentation doesn't explain this stuff adequately.

Tobias Fünke

@SillyGoose oha, nice

@SillyGoose my Weinberg number is 5 and my Preskill number is 4, apparently haha

"the world is a village", we say in german. and even more so the academic world it seems

User1865345

15:01

Hey folks.

I am from CV.

One of the new users have posted kind of not bad answers in a pretty quick manner.

I noticed they also posted some answers here at Physics too. But now I noticed they were temporarily suspended for "plagiarism".

I am not linking the user but if hint is needed:

in Ten fold, 9 hours ago, by User1865345

The nameplay is amazing in the light of the recent folk fandom behind him.

I know mods aren't supposed to talk about why any user is suspended. But now the suspension probably should indicate we would look into their CV posts too.

I am looking into their posts and would talk with the mods of our community, and if there is any info, someone wishes to provide re this, we would be more than happy.

Tobias Fünke

@User1865345 just check this chat here, a few hours ago

User1865345

Okay @TobiasFünke 👍🏻

ACuriousMind

@User1865345 I have raised a flag on one of their CV posts. The CV mods know how to contact other mods if they need more info.

User1865345

Thanks @ACuriousMind.

@TobiasFünke ahh. Nice find. Yes. Same thing is happening at two of their CV posts.

Pretty extensive typesetting. Repetition of redundant facts. I see where it is going.

Appreciate the response and action taken here, folks. 👍🏻

imbAF

15:20

I was reading about the 2 to 2 process and the Mandelstam variables, and the following was said:
Altogether, this gives us 4 × 4 − 8 = 8 independent momentum variables, and the number of independent Lorentz-invariant combinations of these variables is only 8 − 6 = 2.

Where do the 4x4 and 8 come from?
And then the 6?

This was just a statement without explanation

Madder

15:41

looking at the partition function, one can understand that states with higher energy (btw, do we always refer to the microstates when we talk about Z? it is a sum over states) are less likely. But why is this exactly true? If i have a gas, and it has a higher energy per particle, why would this be an unlikely scenario?

Tobias Fünke

@Madder yes, it is about microstates. it is unlikely only because you want the probability distribution to have a certain fixed average value of the energy (canonical ensemble)

PM 2Ring

15:59

Our troll Asim is back with a new colourful name. physics.stackexchange.com/q/837236/123208 Please flag as rude / abusive

Tobias Fünke

perhaps related to your question: physics.stackexchange.com/questions/740370/…

@PM2Ring thanks for noting. But to be honest, I don't think it is wise to engage with them/leave comments etc.

don't feed the troll

Madder

@TobiasFünke i still do not understand the reasoning. this does not seem like a valid physical reason, much more imposting maths to fullfill one desires?

PM 2Ring

@TobiasFünke I don't normally comment to them

Tobias Fünke

@Madder the "valid reason" is the maximum entropy principle.

ok

PM 2Ring

But this guy is very persistent. A comment isn't going to make much difference. He's been trolling multiple sites for a week. I've lost count of how many accounts he's created.

Tobias Fünke

16:08

oh boy...

PM 2Ring

We just need 1 or 2 more flags to auto-delete that question...

Tobias Fünke

I did already

PM 2Ring

Thanks

Tobias Fünke

16:22

@Madder itp.uni-frankfurt.de/~gros/Vorlesungen/TD/… see section 9.1.1. and the preceding derivation/discussion of the canonical ensemble.

Silly Goose

19:19

i am done with classical EM forever >:D (kind of). i finished my final lol.

Arjun

@SillyGoose Yay! I'm going to have Electromagnetism this sem : )

Any tips?

Relativisticcucumber

20:18

@Arjun start EARLY

or else u will end up like us spending 12 hours a day in the local coffee shop for a week straight

a final thank u to @naturallyInconsistent and @TobiasFünke -- our discussions of em waves in materials paid off immensely yippee

Herr Feinmann

@SillyGoose you are never done with EM

Even classical EM is not done with itself

@Arjun yeah, 1A is a lot of current and it could kill you so beware when you try to catch lightnings or something to do experiments at home

Arjun

20:36

@HerrFeinmann Fortunately I happen to have interests lie exclusively in the theoretical aspects of electricity lol!

Silly Goose

20:50

@Arjun Lol I agree with the cucumber: start early. Are you an undergrad?

@HerrFeinmann i would like to forever stay in the quantum realm...and never compute an $\vec{E} \times \vec{B}$ again...

Tobias Fünke

21:06

@SillyGoose congratz!

@Relativisticcucumber sure, welcome-- but I did not do much. Did you have an exam today? or what?

Herr Feinmann

@SillyGoose were it for me I would go back to classical things

Damn Feynman diagrams

Tobias Fünke

well, just don't do QFT :P

ACuriousMind

@HerrFeinmann Once we have bitten the quantum apple, our loss of innocence is permanent

Herr Feinmann

@ACuriousMind I don't even need to check your profile to know you have that quote in your bio :P

@TobiasFünke I'm 150 years late, my guy :P

Tobias Fünke

:d

in which field do you want to work? have I asked this before?

if so, sorry. then I forgot :(

Silly Goose

21:43

@HerrFeinmann you can still have feynman diagrams in classical field theory right :D

Ryder Rude

i like QM too

Herr Feinmann

@SillyGoose tree diagrams are cute

@TobiasFünke Prime Mr. Feynman wanted to work in hep-th

Ryder Rude

There r techniques to do qft with only tree diagrams

Herr Feinmann

The effective action

Now I just want to... I don't know. Do less, get more?

Ryder Rude

4

A: Can QFT not be done with only tree diagrams?

In the path integral formalism, the loop-expansion is the same as the $\hbar$-expansion if the action $S$ does not depend explicitly on $\hbar$, cf. e.g. my Phys.SE answer here. In other words, the classical theory ($\hbar=0$) corresponds to tree-level in the path integral. The operator formali...

the question body is slightly incorrect. i have figured some stuff out since then

Relativisticcucumber

21:52

@TobiasFünke yes! the final

how should i interpret this "____" for the field associated w electrons?

Herr Feinmann

As bullshit :P

Relativisticcucumber

LOL

ACuriousMind

don't you know the electron is the excitation of the ---- field?

----ing electrons, man

Herr Feinmann

I don't really know the context. Maybe they're in a context in which electron is treated as "matter" and only the EM field is quantum

Relativisticcucumber

@ACuriousMind xD

Herr Feinmann

21:55

Like non relativistic QED

Silly Goose

@Relativisticcucumber what book is this from?

kittel

?

Relativisticcucumber

yesh

i was moving away from A&M to get a more conceptual overview but the first thing i saw i was disturbed by

which is that image

Silly Goose

i would also guess what Herr F said. Probably not doing second quantization and also not treating the electrons as classical Grassmann-valued fields

Herr Feinmann

Call me Mr. Feynman please :P

2

Ryder Rude

what do u think is the ontology of particles within our best understanding

the ontology. not just mathematical definition

like, what a particle really is

i think our best guess says that a particle is a concentrated bundle of energy

oh but it need not be concentrated to be called a particle

it can be a global state

Relativisticcucumber

22:03

@SillyGoose what is Grassmann-valued

ACuriousMind

@Relativisticcucumber if you haven't yet seen the treatment of spin-1/2 fields in QFT via Graßmann numbers, don't worry about it

Herr Feinmann

Physics is about going through "don't worry about it", each time at a higher level lol

6

Silly Goose

@ACuriousMind Are grassmann-valued fields also used in QFT?

ACuriousMind

where else?

Silly Goose

classical field theory

ACuriousMind

22:16

but why would you use them in classical field theory if not to "prime" that theory for quantization?

Silly Goose

Okay I see

Hm but are they necessary or just a mathematical convenience?

For a free fermionic field I would imagine we can encode the appropriate anticommutativity via $\{b,b^\dagger\} = 1$ (heuristic expression)

ACuriousMind

@SillyGoose the main thing is that if you want to express the path integral of fermionic field properly you need to make them Graßmann-valued

if you stay purely in the operator formalism you can avoid talking about the concept

Silly Goose

oh okay

Ryder Rude

the Fermionic operator anti commutes with itself. so it can only have Grassman eigenvalues

in an answer, Qmechanic speaks about extending the Hilbert space

cuz the usual Hilbert space doesn't allow for Grassman eigenvalues

But one can formulate the path integral anyway without extending the Hilbert space

think_meaning_buildß

@HerrFeinmann what does that imply about Mathematics?

Ryder Rude

22:25

3

A: Are Grassmann numbers always "under the hood" if we deal with fermionic ladder / field operators?

Supernumbers are needed & used in (among other things): the path/functional integral formalism of a QFT with fermionic fields, cf. e.g. this Phys.SE post. fermionic coherent states, cf. e.g. this Phys.SE post. the fermionic version of the theorem by Stone & von Neumann, cf. e.g. this Phys.SE p...

it says Grassmann numbers r not needed in the operator formalism

because we can do everything with complex Hilbert space

but one can implement Grassmann numbers if one wants to

2

A: Where do fermionic coherent states live?

Consider a fermionic Fock space ${\cal F}$, which is a $\mathbb{Z}_2$-graded $\mathbb{C}$-Hilbert space. Example: Given a single Grassmann-odd creation operator $a^{\dagger}$, the fermionic Fock space is $${\cal F}~=~\mathbb{C}|\Omega\rangle\oplus\mathbb{C}a^{\dagger}|\Omega\rangle. $$ The most...

Transcript for