UmbralRaptor changed the topic of #kspacademia to: https://gist.github.com/pdn4kd/164b9b85435d87afbec0c3a7e69d3e6d | Dogs are cats. Spiders are cat interferometers. | Космизм сегодня! | Document well, for tomorrow you may get mauled by a ネコバス. | <UmbralRaptor> … one of the other grad students just compared me to nomal O_o | <ferram4> I shall beat my problems to death with an engineer.
<kmath>
<BabelStone> Apparently none of my Emoji proposals for Mammoth, Dodo and Ostrich are "well-formed", so Emoji Subcommittee won't… https://t.co/PhR0pSbuY7
<kmath>
<Spectracular> Here's a little Bruker au program to convert dB power levels to Watt. Particularly useful for older systems where o… https://t.co/tCGYFbtz2w
<UmbralRaptor>
Why is -60 dB full power?
<SnoopJeDi>
oh I have no idea. because NMR I guess
<egg>
s/hohhot/ᠬᠥᠬᠡᠬᠣᠲᠠ/
<Qboid>
egg meant to say: /me stares at the ᠬᠥᠬᠡᠬᠣᠲᠠ meeting minutes
<UmbralRaptor>
er, wait. algebra error. >_<
<SnoopJeDi>
#i**i
<UmbralRaptor>
-6
<UmbralRaptor>
o_O
<bofh>
-6dBm is about a quarter-milliwatt and a shitload of power for an NMR
<kmath>
<diodelass> Power brick is a compact (read: refrigerator-sized) free-electron laser producing millimeter waves. Cable is a heav… https://t.co/jhxrUfJ2zp
<UmbralRaptor>
Hm. For some reason, I think of 5 kV as medium voltage instead of SHV.
* UmbralRaptor
stares at power plants.
* UmbralRaptor
likes the idea of power going from a cable tray to a laptop.
icefire has quit [Read error: Connection reset by peer]
<kmath>
<diodelass> Power brick contains a compact synchrotron, which sends electrons down a somewhat-flexible tube lined with confinem… https://t.co/hD6lM7Ggms
<SnoopJeDi>
wouldn't be a synchrotron but that's baaaasically an ERL
<SnoopJeDi>
!acr -add:ERL Energy Recovery Linac
<Qboid>
SnoopJeDi: I added the explanation for this acronym.
<Ellied>
well the laptop end would presumably be one of those
<Ellied>
the power brick would contain a boost ring to get them up to speed before sending them down the "cable"
<SnoopJeDi>
OH LOL
<SnoopJeDi>
lmao a 100 MeV synchrotron you are a monster
<Ellied>
well I mean the one near me is 7 GeV, that's only moderately horrifying
<SnoopJeDi>
(the bending radius for 100 MeV is ~ 3 cm)
<SnoopJeDi>
yea I'm just imagining a teeny tiny synchrotron
<Ellied>
could probably fit it inside a refrigerator-sized power brick given current-slightly-future tech
<SnoopJeDi>
not sure if more or less horrifying than some sort of accelerator-on-a-chip scheme where you pump the chip with MW+ of laser power on each end
<Ellied>
ooh, USB, except instead of positive and negative voltages on metal conductors, it uses two confinement tubes where an accelerator sends a bunch of electrons down one and a bunch of positrons down the other for a one, and vice versa for a zero.
<SnoopJeDi>
pictured: a bold artist explores different media
<SnoopJeDi>
also LOL
<Ellied>
keeps the charge neutral at each endpoint!
<SnoopJeDi>
it's a (CP) symmetric cable, too
<Ellied>
*really* intense twisted-pair cable
<SnoopJeDi>
Calabi-Yau twisted-pair?
APlayer has joined #kspacademia
<Ellied>
best of all, it's easily synchronized on the receiving end with an APD to watch for huge bursts of 511 keV x-rays.
<UmbralRaptor>
Presumably that would make it easy to string the cable wherever.
<Ellied>
Use it for networking and call it IP over Annihilation
<SnoopJeDi>
IPoAAAAAAAAAA
<UmbralRaptor>
IPoγγ
<Ellied>
this sort of sounds like what you might end up with if you had a powerful AI setting up a communication system before anyone managed to explain to it that you don't actually need the same electrons to end up at the other end for it to work
<UmbralRaptor>
Hah
Qboid was kicked from #kspacademia by *status [You have been disconnected from the IRC server]
Qboid has joined #kspacademia
Qboid was kicked from #kspacademia by *status [You have been disconnected from the IRC server]
egg|phone|egg has quit [Ping timeout: 200 seconds]
<egg|cell|egg>
!Wpn whitequark
* Qboid
gives whitequark a barium cyclotron
<FluffyFoxeh>
!wpn
* Qboid
gives FluffyFoxeh a trigonometric stinger
<rqou>
egg|zzz|egg: are you available to help me "debug" some math?
<rqou>
i have a midterm tomorrow, and i have no idea how they're getting what they're getting in the practice problems
<rqou>
and this is impossible to really search for because afaict only a small group of "fancy schools" are using this particular notation/system
egg|phone|egg has joined #kspacademia
egg|cell|egg has quit [Read error: Connection reset by peer]
egg|cell|egg has joined #kspacademia
egg|mobile|egg has joined #kspacademia
egg|phone|egg has quit [Ping timeout: 186 seconds]
egg|cell|egg has quit [Read error: Connection reset by peer]
egg|phone|egg has joined #kspacademia
egg|mobile|egg has quit [Read error: Connection reset by peer]
egg|cell|egg has joined #kspacademia
egg|phone|egg has quit [Read error: Connection reset by peer]
egg|phone|egg has joined #kspacademia
egg|cell|egg has quit [Read error: -0x1: UNKNOWN ERROR CODE (0001)]
<egg|work|egg>
rqou: hmm
<egg|work|egg>
rqou: at work, but ask, I'll see whether I have anything interesting to say
<rqou>
so you remember our evil 6x1 "vectors?"
<egg|work|egg>
hm
<rqou>
they defined a 6x6 matrix that they're calling the "adjoint" matrix
<rqou>
which takes in homogeneous transformation
<rqou>
and gives out {{R, p^R},{0,R}}
<rqou>
(which each thing here is a 3x3 block)
<rqou>
and p^ is the "turn a 3x1 vector into a 3x3 skew-symmetric-matrix" operator
<rqou>
anyways, you can take our evil 6x1 "twist" vectors
<rqou>
and multiply by the appropriate "adjoint" matrices to get a set of new 6x1 "twist" vectors
<rqou>
which is then assembled into a matrix called the "jacobian"
<rqou>
(assembled by concatenating)
* egg|work|egg
has no idea
<rqou>
anyways, they have two "jacobian" matrices, the "spatial" and "body" jacobians
<egg|work|egg>
if that is not explained with vastly more context in the lecture, I'm not sure I'm going to be able to invent said context
<rqou>
it's not
<rqou>
i really have no idea what is happening here
<rqou>
anyways, there is an equation that i can copypasta that is supposed to let me compute these "jacobian" matrices
<rqou>
and it works on the "spatial" one
<rqou>
but no matter what i do it does not work on the "body" one
<rqou>
i even have the solutions, and i cannot get the answer in the solutions
<rqou>
when i asked for help, i more-or-less got "yeah, it should work, try computing harder"
<rqou>
but i can't get it to work
<egg|work|egg>
uh so maybe by spending time on this (time which aiui you can't afford with the looming exam) I could figure out what is conceptually happening, but that's less likely to work if I don't know what computations you ought to be doing, and I'm not going to figure that one out of thin air...
<rqou>
well, there's the textbook "A Mathematical Introduction to Robotic Manipulation" by Murray, Li, and Sastry
<rqou>
but i'm looking through it, and i can't find anywhere that actually _shows_ how to compute the body jacobian matrix
<rqou>
it explains that you can, but when i try it it just doesn't work
<egg|work|egg>
so I have no idea what's going on here, how would I know what specific computation you need to do
<egg|work|egg>
what I could help with is, given all the computations that ought to be done, figure out what's going on; but I can't produce the former ex nihilo
<egg|work|egg>
!wpn rqou
* Qboid
gives rqou an invalid bifurcation
<rqou>
yeah, i think i'm probably going to give up on the "body jacobian" for now
<rqou>
i just have zero insight whatsoever into what's supposed to happen
ferram4_ has joined #kspacademia
ferram4 has quit [Read error: Connection reset by peer]
<rqou>
egg|work|egg: so a much simpler question
<rqou>
if i have a product of two matrices in SE(3)
<rqou>
it seems that i should _not_ be able to convert into the lie algebra and then add the vectors and then convert back
<rqou>
but why not?
<egg|work|egg>
rqou: um
<rqou>
because matrix multiplication isn't commutative
<egg|work|egg>
yeah
<rqou>
but vector addition is
<rqou>
so something is obviously wrong
<rqou>
but what?
<egg|work|egg>
I'm not sure what's wrong
<egg|work|egg>
your statement is correct, but I don't understand the question
<Qboid>
1d 0h 0m 0s left to event #14: JPSS 1/Delta II 7920-10C [at 2017-11-10 09:47:03]. Say '!kountdown 14' for details
<rqou>
if i start with e^A * e^B, why can't i say that that equals e^{A+B}=e^{B+A}=e^B * e^A
<egg|work|egg>
rqou: because that's not how the exponential works on a noncommutative group
<egg|work|egg>
(exp X) (exp Y) is not X + Y, there is an infinite series of nested commutators
<egg|work|egg>
the first-order term being half [X, Y]
<egg|work|egg>
s/is not .*,/is not exp(X + Y)/
<Qboid>
egg|work|egg meant to say: (exp X) (exp Y) is not exp(X + Y) there is an infinite series of nested commutators
<rqou>
hrm, but i recall abusing stuff like this for signals and systems?
<egg|work|egg>
well, maybe if it commutes enough :-p
ferram4_ has quit [Read error: Connection reset by peer]
ferram4_ has joined #kspacademia
<rqou>
ooh that makes sense
<rqou>
iirc when this was abused the matrix was diagonalizable
<rqou>
and diagonal matrices commute
<egg|work|egg>
to quote Trubowitz, "You write commutators to see the level of how much it doesn't commute"
<egg|work|egg>
cc bofh
<rqou>
damn my math sucks
<egg|work|egg>
meow
<rqou>
ok, some of the material is starting to make sense, but i just have no insight whatsoever what this "adjoint" matrix does or how it works
<rqou>
egg|work|egg: alright, maybe you can help with this?
<rqou>
suppose i have an evil 6x1 vector
<rqou>
somehow representing an element of se(3) (the Lie algebra)
<rqou>
and suppose the top half of the evil 6x1 vector is the "translation" part, and the bottom half is the "rotation" part
<rqou>
(so the bottom half is the part that you told me should be a 'bivector')
<rqou>
now i have this 6x6 matrix called the "adjoint"
<rqou>
the adjoint is generated from an element of SE(3) (the Lie group) represented as a 4x4 matrix
<rqou>
and the adjoint is {{R, p^R},{0,R}} like i mentioned above
<rqou>
what is happening here?
<rqou>
this 6x6 matrix gets multiplied to the evil 6x1 vectors
<egg|work|egg>
hmm
<egg|work|egg>
mumble mumble tangent bundle
<egg|work|egg>
there's an element of the group and an element of the algebra, which is a tangent vector
<egg|work|egg>
sounds like this is like our RigidMotion
<egg|work|egg>
rqou: so it might (depending on what the calculations say) be the matrix that takes some sort of degrees of freedom and puts them in another frame that rotates and moves in fancy ways with respect to the first
<egg|work|egg>
the element of se(3) describes the movement (velocity + angular velocity)
<egg|work|egg>
the element of SE(3) describes the current transformation of positions (that's what SE(3) is)
<rqou>
so it seems to somehow be used for _both_ changing coordinate representations and "actually moving"
<egg|work|egg>
yeah that's the usual problem
<egg|work|egg>
any transformation can do that
<egg|work|egg>
and people are rarely strict or explicit about it
<egg|work|egg>
tis annoying
<rqou>
hmm, so if i just assume "it's both" let me see if the slides still make sense...
<rqou>
one part is using the adjoint matrix to change coordinates of "velocity" vectors (which are 6x1)
<rqou>
so that seems to make sense
<rqou>
and another part is using it to "move" "twist" vectors (which are also 6x1)
<rqou>
so i guess that checks out
<rqou>
egg|work|egg: so is this "6x6 matrix * 6x1 vector" essentially doing the same thing as "4x4 matrix * another 4x4 matrix"?
<egg|work|egg>
uh
<egg|work|egg>
whence 4
<rqou>
SE(3) can be represented by 4x4 matrices, right?
<egg|work|egg>
also what does your * mean here
<rqou>
matrix multiply
<egg|work|egg>
ah right, SE(3) has this representation
<egg|work|egg>
but what's the "another 4x4 matrix here"
<rqou>
so the crazy way we tend to do it is that the second 4x4 matrix is an "initial configuration"
<egg|work|egg>
this is a matrix that wraps 1. a SE(3) action on 3d affine space, 2. *magic* on the tangent space
<rqou>
and the first 4x4 matrix is a "change that happens to the initial configuration thanks to the robot arm moving"
<egg|work|egg>
uh
<egg|work|egg>
I don't know, I don't think this has to do with that
<egg|work|egg>
or I don't immediately see it at least
<rqou>
well, i might be wrong
<egg|work|egg>
you have a 3d affine space which we will call "positions"
<egg|work|egg>
you have its tangent space at any point which we will call "velocities"
<egg|work|egg>
elements of the tangent bundle (position, velocity) we shall call "degrees of freedom"
<egg|work|egg>
SE(3) acts on positions
<egg|work|egg>
its Lie algebra is se(3)
<egg|work|egg>
given an element of SE(3) and an element of se(3), you get an action on the tangent bundle
<egg|work|egg>
I think
<egg|work|egg>
maybe
* egg|work|egg
goes back to staring at rtl stuff for a moment, bbl
<rqou>
eh, i'm probably going to give up for now
<rqou>
it's 2:30 am and i need some sleep
<rqou>
if you have any last-minute hints, i'll read them tomorrow morning
<rqou>
at some point i would really like to sit down and actually learn this
<rqou>
btw, now i want robot arms in principia :P
<egg|work|egg>
!u ٪
<Qboid>
U+066A ARABIC PERCENT SIGN (٪)
egg|cell|egg has joined #kspacademia
egg|phone|egg has quit [Read error: Connection reset by peer]
egg|phone|egg has joined #kspacademia
egg|cell|egg has quit [Read error: Connection reset by peer]
<SnoopJeDi>
stream is struggle-bussing a bit on and off again though :(
<bofh>
yaey technology
<SnoopJeDi>
computers: literally impossible?
<bofh>
so like if you were to determine that based on likelyhood of connecting a laptop to a display projector, then yes, the evidence *does* 100% support that conclusion.
<SnoopJeDi>
oh *no* the stream is now left-channel only
egg|phone|egg has joined #kspacademia
egg|cell|egg has quit [Read error: Connection reset by peer]
egg|cell|egg has joined #kspacademia
egg|phone|egg has quit [Read error: Connection reset by peer]
BPlayer has joined #kspacademia
egg|phone|egg has joined #kspacademia
egg|cell|egg has quit [Read error: Connection reset by peer]
egg|cell|egg has joined #kspacademia
egg|phone|egg has quit [Read error: Connection reset by peer]
<Sarbian>
UmbralRaptor: glad to see at least 1 person reads my rants :D
<UmbralRaptor>
Sarbian: uh, somewhat. My limited French knowledge doesn't help. >_>;;
<SnoopJeDi>
!wpn -add:wpn otter
<Qboid>
SnoopJeDi: Weapon already added!
<SnoopJeDi>
good. good.
<UmbralRaptor>
!wpn SnoopJeDi
* Qboid
gives SnoopJeDi a Eulerian spectral FPGA which vaguely resembles a histomorphism
<SnoopJeDi>
ah, a spectral FPGA. This should fetch a high price in the astrobazaar
<rqou>
whelp, totally bombed that midterm
<UmbralRaptor>
Did the Norden help?
egg|phone|egg has joined #kspacademia
egg|cell|egg has quit [Read error: Connection reset by peer]
egg|phone|egg has quit [Read error: Connection reset by peer]
egg|phone|egg has joined #kspacademia
<rqou>
egg|work|egg: can you help me with something else that's you're probably much more familiar with?
APlayer has quit [Read error: No route to host]
<rqou>
i'm trying to get intuition for how the "functions of a matrix can be written using a polynomial because *mumbles a lot* cayley-hamilton theorem"
<UmbralRaptor>
;8ball Does the fact that I'm working with a Hamiltonian mean I can simply assume p*(rxB) is an inner product?
<kmath>
UmbralRaptor: Don't count on it
<egg|zzz|egg>
!wpn rqou
* Qboid
gives rqou a Kolmogorov egg|egg|egg/brioche hybrid
<egg|zzz|egg>
!wpn UmbralRaptor
* Qboid
gives UmbralRaptor a ham Ibuprofen
<rqou>
!wpn egg|zzz|egg
* Qboid
gives egg|zzz|egg an integrand/tern hybrid
<UmbralRaptor>
!wpn egg|zzz|egg
* Qboid
gives egg|zzz|egg an adiabatic polar torpedo
<rqou>
egg|zzz|egg: can you help me with intuition for how the "functions of a matrix can be written using a polynomial because *mumbles a lot* cayley-hamilton theorem" works?
<egg|zzz|egg>
I'm not sure what you're referring to
<egg|zzz|egg>
can you mumble less
<egg|zzz|egg>
also qualify "functions"
<rqou>
"let fhat(s) be any function of s analytic on the spectrum of A and phat(s) be a polynomial such that fhat^k(lambda_e)=phat^k(lambda_e) then fhat(A)=phat(A)"
<rqou>
where 0 <= k <= multiplicity of the eigenvalue in the characteristic polynomial i think
<rqou>
and 1 <= e <= number of eigenvalues
APlayer has joined #kspacademia
<rqou>
"in fact, if m := sum of all eigenvalue multiplicities then phat(s)=a_1 s^{m-1}+a_2 s^{m-2}+...+a_m s^0"
<rqou>
but somehow there's an algorithm that we try and use
<egg|zzz|egg>
well then wikipedia seems clearer than what you're trying to say, try having a look at that
<rqou>
so the algorithm definitely does not look the same, but reading...
<egg|zzz|egg>
that would mostly be because you're not looking at an algorithm? but there are some linsk
<egg|zzz|egg>
s/sk/ks
<Qboid>
egg|zzz|egg meant to say: that would mostly be because you're not looking at an algorithm? but there are some links
<rqou>
well, we were also given an algorithm
<rqou>
but the professor also just sent out a note saying that the algorithm doesn't always work, so ¯\_(ツ)_/¯
<egg|zzz|egg>
there are some links in that section of that article
<rqou>
hmm, i think i understand
<rqou>
the wikipedia article, but not the algorithm we went over in class
<SnoopJeDi>
you can safely assume most algorithms don't always "work" because TANSTAAFL
<rqou>
but the algorithm we were using is _weird_
<rqou>
we write fhat(A)=sum_{l=1}^{number of eigenvalues} sum_{k=0}^{multiplicity of eigenvalue - 1} p_{kl}(A) f^k(lambda_e)
<rqou>
and then we "guess" some f functions until we get enough equations to solve for the p_{kl} polynomials
<rqou>
apparently this algorithm is useful because it allows for solving for the p_{kl} polynomials independent of the function f
<rqou>
p_{kl} only depends on the structure of the matrix A
<rqou>
i guess it would be neat if it worked :P
<rqou>
egg|zzz|egg: so if i'm understanding everything correctly, i can just write out the polynomial the dumb "no neat tricks" way and interpolate it
<rqou>
and this will get me the correct answer on the homework
<rqou>
right?
<rqou>
well, it appears to have worked
<rqou>
i did it the wikipedia way and i got (an invertible matrix of numbers) * (f and derivatives evaluated at the eigenvalues)
<UmbralRaptor>
APlayer: Outside of hardening second strike ability, imperfect (ie: plausible) missile defense decreases stability, though. And in the US case, protecting land based ICBMs might not be needed because of our SLBM fleet.
stark has joined #kspacademia
APlayer has quit [Ping timeout: 383 seconds]
ferram4_ has quit [Read error: Connection reset by peer]
<kmath>
<_vajra> TIL that the kunddaliya, “෴”, is actually in unicode (u+0df4)! Now I really want to start using it as punctuation.… https://t.co/1OBhMAHp1A
<egg|zzz|egg>
!u ෴
<Qboid>
U+0DF4 SINHALA PUNCTUATION KUNDDALIYA (෴)
<egg|zzz|egg>
!kd edit:14 time 2017-11-14T09:47:03Z
<Qboid>
egg|zzz|egg: Invalid ID!
<egg|zzz|egg>
!kd -edit:14 time 2017-11-14T09:47:03Z
<Qboid>
egg|zzz|egg: Updated event #14: JPSS 1/Delta II 7920-10C - A Delta II 7920-10C will launch the BCP-2000-based JPSS 1 (http://space.skyrocket.de/doc_sdat/npp.htm) from Vandenberg SLC-2W into a heliosynchronous orbit - 2017-11-14 09:47:03
<Qboid>
egg|zzz|egg: Description: A Delta II 7920-10C will launch the BCP-2000-based JPSS 1 (http://space.skyrocket.de/doc_sdat/npp.htm) from Vandenberg SLC-2W into a heliosynchronous orbit
<egg|zzz|egg>
should add the antares
<APlayer>
Is there no sort of online database API for rocket launches?
<rqou>
it's called Qboid :P :P
<APlayer>
I mean for use by Qboid, not to have to do it by hand
<rqou>
egg|zzz|egg: i've got some fun abuse of notation for you :P
<rqou>
sup{... stuff, doesn't matter...} = ∞
<Iskierka>
Ellied, I believe 4.4 as it's LTS
<egg|zzz|egg>
rqou: why is that abusive? it depends on the stuff, which does matter
<Iskierka>
I imagine it'll be 4.1* in 18.04
<rqou>
ok, stuff is an integral
<rqou>
the "= ∞" part is a little bit cheating
<egg|zzz|egg>
well that's what suprema do
<egg|zzz|egg>
sup([69, ∞[) = ∞
<soundnfury>
rqou: I dunno, 'sup with you?
<egg|zzz|egg>
that's a perfectly nice supremum
<rqou>
is it actually correct to say that it equals infinity though?
<rqou>
i thought that's not quite correct?
<soundnfury>
rqou: depends whether you're working in the reals or an affine compactification no?
<egg|zzz|egg>
also sup(∅) = -∞
<rqou>
wait really?
<rqou>
i guess that works
<soundnfury>
has to be a two-point compactification, of course, the one-point isn't ordered
<rqou>
i don't know what a compactification is
<soundnfury>
rqou: it's when you adjoin extra elements to your field to make it compact
<soundnfury>
in the case of the reals, you can either do the one-point compactification, where you add an unsigned infinity (giving you a somewhat circle-like topology)
<soundnfury>
or the two-point compactification, where you add positive and negative infinities
APlayer has quit [Ping timeout: 383 seconds]
<egg|zzz|egg>
!wpn Iskierka
* Qboid
gives Iskierka a chromium char conotoxin
<egg|zzz|egg>
!wpn UmbralRaptor
* Qboid
gives UmbralRaptor a conjugate clowder
* soundnfury
is kinda disappointed that egg didn't immediately tell me I was wrong somehow
<soundnfury>
!wpn egg|zzz|egg
* Qboid
gives egg|zzz|egg a Norman classmethod
awang has quit [Ping timeout: 383 seconds]
<UmbralRaptor>
egg|zzz|egg: Algebra is a Lie.
<SnoopJeDi>
bofh, just attended a lecture by Sachdev, SYK is *neato*
<SnoopJeDi>
also, he presented an AdS/CFT correspondence idea of SYK in 2+1 being holographically matched to Einstein-Maxwell description of a charged black hole, and drew a question from Nick Suntzeff about why that's practical because anti-de Sitter space isn't very universey.
<SnoopJeDi>
Sachdev deftly answered "I'm a condensed matter theorist, I don't really care about the universe" and got a room full of laughter and applause (he extended his answer to something a bit less cheeky)
<bofh>
SnoopJeDi: holy fuck, Sachdev is amazing
<SnoopJeDi>
HE IS
<SnoopJeDi>
Easily the best CM talk I've ever attended
<SnoopJeDi>
and I could even *follow* it and learn something
<bofh>
I've talked to him before at conferences, the sort of person that's nice but blows you away with knowledge in like, every sentence.
<bofh>
YEP
<bofh>
His talks are great like that. He knows how to aim them correctly for the audience.
<bofh>
He also wrote a very good CMT textbook btw.
<bofh>
also I love his response. It's so true tho.
<SnoopJeDi>
Quantum Phase Transitions?
<SnoopJeDi>
That one got a box when I realized how frickin' cool his 2+1 correlated electron model is in contrast to the uncorrelated result
<bofh>
Yep, that one.
<bofh>
I've used it partly for one of my courses.
<SnoopJeDi>
although I don't fully grok how you formally prove the model doesn't support quasiparticlces
<SnoopJeDi>
do you write out the Hamiltonian expanded in interaction order and somehow prove that the limit of more interactions doesn't converge to the known result or something?
<SnoopJeDi>
he presented a visualization of the energy levels and argued that with quasiparticles, the low energy states are spaced by ~ 1/N, but in the SYK case were spaced ~ exp(-N), therefore obviously no quasiparticles
<bofh>
iirc more or less, tho that limit is kinda nasty.
<SnoopJeDi>
I imagine, considering the work got going in the 90s, but he's published a paper in the last year on it
<SnoopJeDi>
although it seems like recent stuff is maybe more concerned with AdS/CFT correspondence than with "strange metals" as he calls them
<bofh>
So the thing to realize is that the current batch of CMT work is based on ideas formulated in the mid-90s and supported by experiments from the early-to-mid 2000s.
<SnoopJeDi>
s'fair
<bofh>
It's actually a great time to be a Condensed Matter Theorist, funding/govt insanity aside.
<SnoopJeDi>
I didn't mean that as a dig, heh
<SnoopJeDi>
yea we're heavy on CM here, very well attended talk
<bofh>
I figured you didn't, just sorta providing background as to the state of the field.
<SnoopJeDi>
although the non-dept people were rude af during Q&A
<bofh>
fuck's sake
<SnoopJeDi>
lots of chatter and noise as they shuffled out, and then that gets the normally well behaved people going and ugh
<SnoopJeDi>
it's been an hour, sit still for 5 more frickin minutes and you might learn what a good question is
<bofh>
Yeppppppppppppppppppppppp
<bofh>
Doubly so since a lot of the senior non-dept people I find here wouldn't know what a good question is if it hit them over the head.
<bofh>
Dunno if that's true there too but it certainly seems like it.
<kmath>
<kevinmgill> Saturn's moon Iapetus imaged w/ infrared, green, and ultraviolet filters on December 27 2004 by @CassiniSaturn -… https://t.co/tA0izeeu31