« first day (2925 days earlier)      last day (2300 days later) » 
00:00 - 16:0016:00 - 00:00

16:00
Ugh, I am losing patience of you mathgeek
in Mathematics, 1 min ago, by Math geek
Henno should be wrong. am I correct?
Anonymous
@NovaliumCompany Tough question. I'd say partly I do but on the other hand, there was no other better option either
A combination of RHV + "am I correct" + "I am right you are wrong" = recipe for disaster
I still think that strings phd description we found was the best choice of them all
@bolbteppa which one?
@Blue You could have started work? You seem pretty brilliant and smart to me so it shouldn't be hard?
16:01
This is the 3rd time I saw him exhibit this phenomenon
Anonymous
@NovaliumCompany What type of work?
Remember, the big list, I have it bookmarked somewhere
Anonymous
Life is extremely difficult in my country without a basic academic degree. Also the chances of you getting hired is almost nil
also time to head home
brb
thanks
16:02
'Applications of supergravity and string theory to high energy physics'
It's still open
@Blue Yep, I can relate. How about programming?
And now Slereah is in here, I better just get off h bar before I cannot resist the urge to flood and hence been banned for the 4th time
Anonymous
@NovaliumCompany Well, good programming jobs are very competitive. It's way easier to get into those jobs if you have an undergrad degree
Anonymous
Without a degree, it's gonna be a tough and uncertain path ahead
16:04
It's rare you get so many options for which direction you want to go in the description
and it even says in the beginning you'll be given time to read up
@Blue I think most people (atleast in the coding industry) don't hire by degree anymore, they hire by skill. I donno, I just feel like 3 or 4 years of your life is a lot to waste.
Anonymous
> I think most people (at least in the coding industry) don't hire by degree anymore
Anonymous
Oh dear :P
Anonymous
Maybe that's true when you already have say 15+ years of experience already
Anonymous
Not when you're just a 18-19 year old kid
16:06
@Blue Actually, I can understand you. I know India is hard, a lot of competition.. so, I think you did good in entering uni.
@Blue Have you even thought of moving somewhere else?
Anonymous
@NovaliumCompany Yeah, I'll be moving out in a couple of years for grad school
@Blue Grad school? What's that?
Anonymous
@NovaliumCompany Graduate school
@Blue Ph.d?
Anonymous
@NovaliumCompany Depends on the country I guess. In some places a masters degree is mandatory before PhD
16:10
@Blue. Ok, and what about when you finish grad school?
Anonymous
I dunno...lol
Anonymous
I usually don't like to plan too far ahead
@Blue So you are doing all of this academic stuff... grad school... and in the end, you actually don't know how you want to continue?
Anonymous
That's right
@Blue Alright. I feel like university, grad school... are just a way to delay the fight with life and uncertainty, but it will eventually come.
@Blue Nevermind, that's just my point of view.
Anonymous
16:17
I honestly don't think uncertainty is a bad thing. My principles are like greedy algorithms i.e. I only plan for short-term goals...with the hope that in the future I'd still get to do something I enjoy.
@Blue Ok :-)
Is there a name for a function of the form $f(x_1,...,x_n) = f(u)$ where $u = c_1x_1+...+c_nx_n$?
Ugh, not $f(u)$ but $g(u)$ of course
Consider for example a wave of fixed shape traveling in some fixed direction at constant speed $v$ - the displacement function is then of the form $f(x,t) = g(x-vt)$, if we place the $x$-axis in the direction of propagation
16:36
@Slereah Henneuax has said he definitely has no PhD position but only has postdoctoral ones.
@CaptainBohemian No matter
There's plenty of people at that lab
I can just pop in his office to ask him to sign my copy of quantization of gauge systems :p
Wow!What a radical approach!
is Henneuax a nice man?
Not a clue
I feel he even doesn't read my mail.
He can't ignore me if I'm in his office
Anonymous
16:48
@Slereah I hope that's not a challenge :P
Barnich seems to be nicer because he did read my mail.
Geroch's pretty nice too
and Krasnikov
On the other hand
YURTSEVER
You're on the list!
Anonymous
@CaptainBohemian Why would you decide niceness on basis of whether they read emails? :/ Professors can be busy...
Anonymous
I do agree that it sucks to not get a response tho
@Blue I can't tell completely if a person is nice from email communication, but I just guess.
Hanneuax replied me but his reply implied he didn't read my mail in detail.
16:52
@CaptainBohemian UK universities are businesses and foreign students are one of their primary sources of income. What you're describing is just advertising by a business to one of its primary customer pools. I don't know why you find it strange.
hmmm
@JohnRennie Certainly looks to be the case.
But that doesn't make it less true though.
@EmilioPisanty because I seem not to have seen other countries advertising PhD positions which only fund people from certain areas.
Anonymous
@CaptainBohemian Which other countries? Example?
Anonymous
I've seen tons of advertisements for US and Australian universities in my city
Anonymous
16:57
Even some of the Asian unis
Anonymous
(and most of them offer very little or no funding)
sigh wrong chat
All advertisements I see from other countries have no limit of funding area.
Anonymous
It probably has more to do with the contacts the event organizer in your area has
Anonymous
Also, universities only advertise in those regions from where they're more likely to get students
17:01
@Blue if no funding, what bother to advertise? in my country, there is seldom PhD position advertisement because it isn't counted as a job.
Anonymous
> if no funding , what bother to advertise?
Anonymous
9 mins ago, by Emilio Pisanty
@CaptainBohemian UK universities are businesses and foreign students are one of their primary sources of income. What you're describing is just advertising by a business to one of its primary customer pools. I don't know why you find it strange.
@CaptainBohemian because people with money do apply
either they self-fund, or they get third-party funding
which is all the same to the university
if it worries you that this would select for only a very small fraction of rather rich students, then you're not alone in that worry, but it's not a concern shared by the universities.
@JohnRennie damn. There was a great QI qip (or was it Mock The Week) about this, but I can't find it.
@Blue I don't know what you mean by regions there. I read advertisements from web which can be seen everyone in the world, not just my regions.
17:26
@EmilioPisanty actually the uk positions I see on web all have funding (that's why I said if no funding, what bother to advertise?) but if you click the advertisement to read the detail, you will find the funding is only limited to EU-students.
@EmilioPisanty the only QI quip about Germany I remember is this one
man...these gaming laptops are all so shite...
expensive and only with GTX1060's...
psh
17:43
any gaming experts in here :P
Yeah get a desktop :D
There's something nice looking
> They enjoy a decoupling region close to their horizon
How presumptuous. Maybe the black holes hate it!
I already have a desktop
but it runs linux so
gaming is difficult on it
Dual boot! Though I've ran into a number of bootloader issues doing that
17:50
I have always run into bootloader issues doing that lol
every time I've gotten dual boot my computer fails to boot and I have to use something like a grub-fixer thingy
@EmilioPisanty NEIN!
OCTO
This was more along the lines of "what has Merkel given the Germans this week?" "A huge tax break" "so how can they afford it?" "Because they're not broke. Unlike everybody else."
17:53
It's a bit easier if you put each OS on a different drive and just boot to a different drive. Not as convenient as having all the options in a grub menu though
I'm not gonna mess with dual booting on my desktop
You mean you don't want to spend a weekend hitting your head on a wall?!
Wow I was rejected for a job I didn't even apply for...I don't know if that's sad or funny
lol
mostly funny I think
And yeah, I don't want to spend a whole weekend banging my head on a wall
It was actually one a recruiter invited me to apply for. Evidently that somehow automatically applied me to the position whether I wanted it or not
fail
I have a 15 minute chat with openai today so that's cool I guess
Anonymous
18:00
@enumaris Congrats!
:D
I don't know how many more steps are in this process -.-
Anonymous
Is it some interview or just casual chat-with-recruiter?
All of them
they call it an interview
but it's 15 minutes
so I guess it's more like a chat
but it's not with the recruiter, it's with someone else at the company
maybe they thought my solutions were borderline so they put a 15 minute chat to see if they can just filter my application out XD
an optimistic view might be my solutions were so awesome they thought 15 minutes was enough to just make sure I wasn't some sort of psycho or something but otherwise they've already made up their minds
the reality is probably somewhere in between there
18:06
Maybe to figure out where your interests are or something?
well obviously since I applied there, I'm interested...o.O
But yeah maybe they just want some more preliminary info
But they do research in a variety of different areas, right?
yeah I think so
with the overall stated goal of AGI
the recruiter yesterday from Beyond Limits also mentioned some project they are working on to produce "human like intelligence" from some chip
I dunno how much she really knew about that project tho
cus I asked her if it was similar to the google assistant and she said not really
There's also a company called ModuleQ that's working on some sort of "google assistant" like "intelligence" to help with work-related issues...looks like an interesting concept
I think a recruiter is presenting me to them, but I haven't heard back yet
help?
1
Q: What is the source of discontinuity of heat capacity?

More AnonymousHere's something that's been bothering me. Given the density matrix of a system: $$ \rho = \sum_{ij} p_{ij} |\psi_i \rangle \langle \psi_j |$$ Now, we can define it's entropy given by: $$ S= \text{Tr}(\rho \ln\rho) $$ The energy of the system is given by: $$ E =\text{Tr}( \rho H)$$ Hence, ...

Hmmm I feel like I've heard of something like a voice assistant for work. Don't remember where though
18:19
lol
.68 f-score is the top
sad
probably just run an RNN on it
done
.2 f-score
@MoreAnonymous It's not quite the same as what you're asking, but I think it's related: How can the partition function $\mathcal{Z}=\text{tr}( e^{-\beta H})$ ever fail to be an analytic function?
I think the point, if I'm remembering right, is that you only see phase transitions in the thermodynamic limit
vzn
vzn
@enumaris congrats! really wish someone from openAI would show up on SE! btw you might have to develop some opinions about AGI! :P ... just saw this by sutskever from oct 23 whaddya guys think
at which point the presumption that $\mathcal{Z}=\text{tr}(e^{-\beta H})$ must be analytic will break down
18:24
can't view videos at work
@Semiclassical maybe write an answer?
Eh, what I'm saying is too vague for an answer as yet
vzn
vzn
(view this!) holy @#%& sutskever making close to $2M/yr o_O nytimes.com/2018/04/19/technology/…
But iirc what you find in some cases, for instance, is that for a finite system you'll have $\mathcal{Z}$ possessing some set of poles in the complex plane of $\beta$
18:25
and then these poles, in the thermodynamic limit, accumulate into lines which give rise to non-analytic behavior. I think the usual example of this is with Lee-Yang zeros?
the main upshot is that phase transitions (and their accompanying non-analytic behavior) don't truly exist for finite systems
Which should sorta make sense: As you approach a true phase transition at $T=T_c$, you expect to see certain quantities diverging as $|T-T_c|^{-a}$. That'll include stuff like the correlation length
but this only makes sense if your correlation length, while large, is still smaller than your system size
so to get a truly infinite correlation length you need an infinite system
well, that didn't work like I meant it to
@MoreAnonymous maybe read this Q&A:
5
Q: Free energy functions are analytic or non-analytic in phase transitions?

SaavestroI already saw this Phys.SE post and it seems perfectly reasonable that the free energy describing a system must be a non-analytic function in order to display a phase transition. An analytic function can be Taylor expanded around the critical point. In a phase transition the derivatives of the ...

@vzn yeah I've seen that. Unfortunately, I doubt I'll be paid that much lol
vzn
vzn
18:49
@enumaris maybe you could get a serious raise if you get more serious about AGI :P
Just win enough kaggle comps to make that much, duh!
maybe, we'll see
AGI is not happening at my current company tho
that's pretty much out of the question lol
Attention seems very interesting...haven't had to work with it so far tho
vzn
vzn
19:13
@enumaris lol AGI is not "happening" almost anywhere except maybe glimmers )( at deepmind + openAI.
well I mean an initiative to work towards it is not happening here
@enumaris I still have tabs with the papers sitting there to be read lol
vzn
vzn
@enumaris lol not surprising eh? you said its an insurance company :P
vzn
vzn
@danielunderwood it would be really cool if kaggle winners got more exposure. they get some... blog.kaggle.com/category/winners-interviews
19:19
Some of the competitions give finalists the ability to talk at conferences I believe. From what I've heard, winning is more a game of luck though
one would guess that all the models are already known
and ain't nobody coming up with something totally novel like the attention mechanism for a Kaggle comp
so it really just comes down to model and hyperparameter selection and data cleaning
There are quite a few that are won with logistic regression or decision trees I believe
Well forests
vzn
vzn
the top competitors are aware of most models but add unique "polish". do wonder if anyone can win consistently/ crosscuttingly. maybe its not happening.
 
1 hour later…
20:51
whoop, just had a pizza party lol
pizza was not bad
quite full tho...
I sure could go for some pizza
interesting library
apparently they've implemented a bunch of models
and datasets
into one package
they have a IMDB dataset for sentiment analysis training
and they implemented the transformer architecture in it
neat
no need to code a transformer from scratch by hand lol
now you can literally build your own google translate
XD
21:30
interesting
it errors out python when I try to import the models lol
all of the documentation goes through command line level stuff...hmmm
I went to the Academia SE chat for help
But it is a ghost town
ah, I now get an error when I try to load tensorflow
uhhh....
@Slereah Haunted by the ghosts of grad students who didn't get a job? :P
I don't get a traceback....so I don't know what the error is....uh oh...
What does the error say? Surely not just "Error. You lose.".
21:35
@ACuriousMind that's me
Hey any of you jerks are fancy professors?
Hire me plz
@ACuriousMind it's not an error within python
Anonymous
@Slereah Their main site is quite responsive :)
a windows prompt shows up and says python errored out
and python closes
I have not had that happen to me before lol
uhhhh.....
looks like I'm not gonna be trying to install this on the actual big computer that matters lol
or I guess I can set up a vm first and try to install it on the vm
I can just look through the source code to figure things out if worse comes to worse I guess
trying to upgrade tensorflow...but it's frozen after the uninstall part...
uh.......
whoop finished
still errors out......
well then...wtf...
22:01
Sounds a lot like the first time I tried to install tf
python itself errors out?
Oh no I'm just saying the "uh......wtf" part
What's the error it shows?
it's the windows pop up
"python.exe has stopped working..."
there's no traceback
Python has stopped working. A problem caused the program to stop working correctly.
ah, I can click debug and at least see the error code
41932
but it just says "an unhandled win32 exception"
Just fix the problem that causes it to stop working correctly, ez
if not working_correctly:
    fix()
looks like this errors in other packages because those packages have only unix support
but...tensorflow has windows support
22:07
They could be importing one of those packages?
would be dumb to import a package without windows support in a package that is for windows...
unfortunately if the error is cus it's importing "some package" that has no windows support...that's going to be extremely hard to track down...
not really sure what to do atm about this lol
Are you running directly from a command prompt/powershell?
from anaconda prompt
it errors whether I open python and then type in the import lines
or if I do a python -c "import tensorflow"
And I guess it kills the shell where it can't even print anything?
the shell doesn't die
but it doesn't print anything
python just exits
so I'm back to C:\Users\XXXX>
22:22
Try creating a fresh venv and installing just tensorflow or tensorflow-gpu?
Or check to make sure your version of tf is compatible with the version of cuda and cudnn you have
Or could try cpu-only version to see if it crashes
Though if you have the gpu version and try to install the cpu version, it could break things
22:47
right now I'm only using cpu only
0
Q: What's wrong with my question?

4dr14n31t0r Th3 G4m3rThis is my question: Position of a electrical charge with mass at a given time under the influence of the electrical field of another charge It has been put on hold for reasons I can't achieve to understand put on hold as unclear what you're asking by Chair, Aaron Stevens, Kyle Kanos, Acci...

stupid openai dude called me 15 minutes late for a 15 minute call lol
rescheduled for friday
I have tensorflow installed outside a venv so I guess if I create a new venv it will still have access to the tensorflow install
Pretty much every phone interview I've ever had has managed to be late lol
And I think it may depend on a site packages option. Not 100% on that though
Yeah I probably still could have taken the call, but it feels very rude
so rescheduled for Friday meh
23:18
I always feel better when they start out a call by apologizing lol
lol
they are indeed often late...
Meanwhile I'm just staring at the clock for the ones I have to call in for. It seems like all of my recent ones have been them calling me though
Although I'm usually staring at the clock for the ones where they call me too
23:33
mostly they call me
I've called one guy
and that was the dude who spent the whole time telling me I wasn't qualified lol
still broken
hmmm
The obvious solution is to nuke your computer and start fresh
also my task scheduler is not running the task I just scheduled due to "the operator or administrator has refused the request"
hmmm
try again!
seems to be running now
w/e
the scheduled task, not tensorflow...
tensorflow is being annoying
00:00 - 16:0016:00 - 00:00

« first day (2925 days earlier)      last day (2300 days later) »