/sci/ - Science & Math

>>9215659
I did not have any involvement in this project in any way, either academic citation wise or actually being involved wise, after this one paper.

>5 authors

So let me guess, you wrote the draft but didn't do anything

it has a lot of words so i like it
charge your phone

>willingly uncloaking
I'm sure your advisor would LOVE to know you browse 4chan.

>>9215670
I did write 98% of the actual code and all the paper drafts for this specific part of the project, but I was not involved in the individual projects of any of the other authors including Chris Snow himself (except in a academic advising capacity).

>>9215672
See >>9215681

I do realize why you would automatically think that, but it wasn't like that. I had a previous SURF before that for the DARPA urban challenge where I really basically didn't do anything or write any drafts or publish anything.

>>9215689
I mean in that one I was writing Python code for simulated Mars rovers at JPL, which is very cool, but that's just writing 100 lines of code with pre-existing Python libraries, not actually doing something.

>>9215695
Probably wasn't even 100 lines, probably more like 5 lines, I was just including cases where I just edited an already existing line

>>9215702
I did write the "libtrafsim" Python library, but that's not doing something, that's writing a Python library for a team that was already working on simulators using Kalman filters and shit, whatever those are. I'm a CS major and I still don't even know. I was a sophomore at the time.

>>9215678
I like you too
>>9215679
I don't have one, I bailed out of academia entirely with my bachelor's degree and went to Wall Street

>>9215733
You could say I met my match in Chris Snow and realized I could never match his higher native intelligence and more years of being enthusiastic about his particular field than I've even been alive

>>9215740
I am thankful to Caltech for mercilessly exposing me to this un-PC reality before I'd made the mistake of thinking about which grad school I should apply to

>>9215755
This trailer kind of reminds me of what these SURFs were like

Stop bumping your own thread you autistic shit stain, why are you flaunting a paper from 9 years ago

>>9215784
I'm just trying to show off my paper for fuck's sake. Do you know there are actual PhDs on this site?

>>9215814
It's just one little undergrad research project paper. I don't see why any PhDs would be afraid of it. None of the PhDs were afraid of it at the time, so what's the problem?

>>9215826
If there's one person you should be afraid of, it's the collective loli known as the Arnold Lab at Caltech

>>9215835
Not to be confused with the Baker Lab, which is the collective loli from whom I stole all my ideas for the benefit of the Arnold Lab

>>9215842
In conclusion, everyone should contribute spare CPU cycles to the Baker Lab's Folding@Home project.

>>9215845
Aren't academic citations 100x more fun when they're handled by software and you don't have to go up on stage looking like an ass saying "no, *I* did that"?

>>9215857
In a legal sense, it's kind of like double jeopardy. Why would anyone have to defend, much less a PhD thesis, some random paper they've already published, all over again from first principles?

>>9215862
You could even imagine a machine learning AI that defends a random paper from its own citation tree from first principles, but that's a PhD thesis all in its own

>>9215873
I know it has to do only its own citation tree because in the general case for any paper, that would be reducible to some kind of superintelligent science AGI.

>>9215882
If it's only the citation tree, you'd need a full learnable model of academic peer review, but that's not even a PhD level problem

>>9215890
However, for the record, it's also not really an undergraduate level problem, which leaves the poor SURFs in an awkward position

>>9215901
Now Yaser Abu Mostafa is surely having second thoughts about his YouTube CS 156 course, but in what possible alternate universe could I have had any coherent complaint about this course at any time, in the past or foreseeable future?

>>9215910
I know I have been making a lot of incoherent complaints about a lot of thibgs ever since I took this course, but never about the course itself

>>9215913
This also explains my time at Virtu. I had a lot of incoherent complaints in my head about most of the things anyone ever said to me, but I was never able to express them coherently

>>9215928
forgot my pic

>>9215931
Yaser Abu Mostafa said to me, not directly but through his TAs "I'm not sure I understand what part it is you don't understand". I replied, "I don't know either, but here's my Netflix challenge thing with SVMs or whatever".

>>9215944
Much like all my other graded CS courses, I got an A in this class even though my actual individual grades on the homework and finals before the curve were an F-

>>9215958
I remember being a TA in CS 1a and giving all my students an average of 0, which must be an F-- so this is probably fair

>>9215967
In English, an F-- would be, not only did d you fail, but all your friends did too, just not quite as badly as you did

>>9215972
In any other school, this would sound like a serious accusation of cheating, but Caltech has a strict honor code that I have mostly never violated and mostly never seen anyone else violating

>>9215977
How have I violated the honor code? This is a true Caltech confession. I have never collaborated with anyone against the honor code, but on multiple, escalating occasions I have taken more time than I was supposed to, and even looked up theorems and proofs in math papers on the internet. Don't try this at home, kids.

>>9216003
Who knows, they might even have to lower my GPA from an A to a D for this, like it was before the curve, and take away my upper class merit scholarship. Thank God I've already graduated!

>>9216006
But what if it's all the way to an F, and not just a D? That's for Caltech administration to somehow decide. This was like 10 years ago after all.

>>9216019
I knew those 80% of Caltech courses where I was failing hilariously but switched to P/F on drop day and ended up passing had to be good for something

>>9216033
You can tell which courses I cheated on the least by how badly I was failing, except for that fucking intro number theory course where I tried to cheat as hard as I could but still almost failed until I figured out how to prove like one of the theorems

>>9216041
So from a formal perspective, this is one class in which I think (IIRC, if I even passed it and didn't drop it on drop day entirely) where I got a P but definitely deserved an F. But then you'd have to reevaluate all my other Ps as well, and those cases are much less clear

>>9216054
There was no penalty for this behavior in any of my classes at the time, but it was certainly known to be, and turned out to be in actual fact, very bad karma

>>9216073
Fortunately for me, ever since then I have been, at worst, involuntarily psychiatrically hospitalized 12 times by my retarded mother Alfia and my retarded ex Caltech gf Lauren, so how much more can I even expect for cheating a few times in college?

>>9216082
Maybe it's been Richard Feynman's ghost that's been haunting me this whole time, bearing an indelible grudge for my getting an unfair undergrad Caltech grade

>>9216092
In all serious though, it would actually be a grave insult to Richard Feynman himself if nobody corrected both my grade and my hospitalizations.

>>9216103
If Richard Feynman himself were to inspect all my Caltech homework, what do I think he would find? A lot of misguided nonsense in bad handwriting, a few good ideas here and there, and a lot of Ps instead of Fs

>>9216119
Better to have Feynman say to you "this entire line of reasoning is complete nonsense, and here's why", rather than "I literally don't even know what you're talking about in this specific sequence of English words in your proof"

>>9215659

This thread has been quarantined to control radioactive levels of autistic same-phag blogging.

>>9216135
I guess the only way for an idea to be so completely wrong that even an undergrad TA can prove it wrong is for the proof itself to be blatantly self-contradictory. Otherwise you might need to call in a PhD

>>9216151
How can an undergrad write an entire pages long proof that is blatantly self contradictory from line 1 and still feel good about it? It's because he or she is not a PhD

>>9216155
The worst case is when it's absolutely right on lines 1-10, wrong on lines 13, 57, and 102, and you have to explain in the margin in red pen which particular bag of unholy CS or ML theorems actually proves this wrong

>>9216168
Thank God in my SHARPEN paper there was only some kind of jerry rigged type theory and distributed computing stuff and no actual machine learning involved. We didn't do any machine learning at Virtu either the whole time I was there

>>9216179
I was writing in my journal recently about a 2D or 3D vector/video/video based symbolic human readable language based on cons cells and machine learning, but I never got anywhere with this on my own

>>9216188
I was thinking about a language that with backreferences could automatically look things up for me and rewrite things in my own words using machine generated cons graphs, but this was impossible for me. I was using the Python library moinmoin and vim

>>9216194
You can even do local, static, or dynamic time varying cons hypergraphs in N to infinite dimensions. There is no limit to what you can do with category theory, which is also something I don't know and have never studied

>>9216198
Or global, if you dare. But global is only relative to your current physical machine and network topology, process ID, and symbol table

>>9216201
This language would ideally be JIT compiled on the fly as you type, and not interpreted

>>9216216
In SHARPEN, this would probably be written as

> import Sharpen.Distributed.MachineLearning.JIT
instead of
> import Sharpen

With a custom Python import module to resolve the hierarchy of the submodules

>>9216226
There is even, if you are prepared to truly lose all your sanity and reason, an exception that the import module throws if it can't find a module, that you can feed directly back into the machine learning distributed JIT compiler

>>9216236
You could even make an entire hierarchy of multiclass exceptions derived from LoseAllSanityAndReasonException

>>9216243
You could even make an ImNotSureWhichSpecifcSanityOrReasonIHaveLostButImThisSureIAmConfusedAboutTheFollowing(*args, **kwargs) exception to throw from within the exception handler

>>9216250
Which calls into an ordinary external Python/C++ library that parses stdout/stderr back into the machine learning JIT to obtain an estimate of how confused it is about what, specifically

>>9216257
If you need more data than every compiler of every language running in parallel, you can ask the NSA for its back and real time archives of every 4chan post and its associated IPs and post ID / timestamp cross references

>>9216263
Or you could peer into your local hedge fund's 4chan-scraping market data venue

>>9216250
You could also make a IAmNotConfusedAboutAnythingButIAmSadBecauseILostOnPerformanceAndOrAccuracyToTheFollowingOtherWorkUnits(*args, **kwargs) exception, or some multiclass combination of the two. The ML also could even generate the exception hierarchy itself

>>9216280
You could even parametrize the sadness by flavors and intensities and have the exception write an English poem about it on demand, so long as you don't confuse this with actual human emotion

>>9216287
If you really squint, a machine learning model trained on the output of these exceptions as data could probably learn to write an English poem about why it's sad about any given English language subject

>>9215659
Well, I'M fucking interested.

Expatiate.

>>9216293
You could even create a UI for this for sad humans to intellisense style write a poem about their own particular existential despair

>>9216298
Please and thank you.

You are a treasure, and a pleasure.

>>9216299
I call it, Microsoft Poem. I was going to name it Microsoft Sentence after Microsoft Word, but that has some bad implications to it.

This paper actually reminds me of my favorite song: https://www.youtube.com/watch?v=br517ctCUCE

>>9216303
Obviously you can also just change the sentiment flag from sad to happy to change the genre of the poem, but don't stop there. Make sure to also include the "pasta" and "zucchini" sentiments in addition to just happy or sad

>>9216317
If you hook this up to Google or OpenAI's cloud services, we could watch in real time our 50000th technological singularity since 2014

>>9216324
The only time parameter left in that case is global cloud and secret military compute capacity, which can be backed out from US equities and futures market data over the past several years

>>9216332
BR ratio = 1

Welcome to Humanity.

>>9216332
If you can't prove Yudkowsky's "Lob's Theorem" or whatever wrong using normal Aristotelian logic, can you at least prove it wrong using some or all of the alternative logics they have in the foundations of math departments in academia?

>>9216352
At most, I bet you'll be able to at least prove that it can't be proven wrong using Aristotelian logic. I am 99% confident that Yudkowsky is retarded, but not that retarded

>>9216366
If it's that 1% and he really is that retarded, I swear to god I am going to personally kill this guy

>>9216380
Hey asshole, it's a positional-queue.

>>9216384
Me first. Please and thank you.

>>9216380
"Anything follows from a contradiction" is a true statement, but what is this "anything", and what's a "contradiction", really? This is a question for paraconsistent logic or possibility theory or any number of things I've never studied except on Wikipedia

>>9216402
What's "anything"? It's the opposite of "nothing", but what's an "opposite"? I'm just guessing here based on what I know from Wikipedia

>>9216408
An opposite could be just about anything, but what if it's nothing? I've honestly never heard of a logic like this, nor do I expect to

>>9216413
Actually, I've probably heard of it somewhere, but it's probably some null object in category theory with barely any axioms or theorems in it

>>9216419
OK let me go nuts here for a minute - what's "an" axiom anyway? If I channel Hilbert here, an integer scalar number of axioms may only be a special case anyway

>>9216422
Since this is Hilbert we're talking about, even the most fully (?) general case might only be a special case after all. You never know

>>9216430
If you use this line of reasoning to prove the existence of God, make sure you didn't accidentally only prove it in some alternate domain of the multiverse, after which God is forced to send an army of anime characters after you to psychologically and emotionally abuse you

>>9216437
Your JITting AIs should compete to have a worst (by some GR metric), most inane conceivable argument that still happens to be right (by some other (?) GR metric), and see if they can beat this 4chan thread

>>9216443
There's also a Turing test in here somewhere, though I can't precisely find it. If I were a computer, this argument would be either much scarier, or much more inane, or both at the same time in some bundle of mostly orthogonal directions

I don't know about you but all my parents beat me.

>>9216446
Which by the way brings me back to this video, >>9215765, on a mostly unrelated subject

>>9216450
oh yes, except that one. Leisure Suit Larry Sierra. Tango Delta Foxtrot? Woof woof.

>>9216467
IDontCareIfItsEvenCorrectOrSafeIJustWantToEitherInspectOrRunThisCodeException(*args, **kwargs) is probably useful too

>>9216514
WtfButWhateverIJustWantToBeOnThisCitationTreeSomewhere could be a subclass of that for the peer review model

>>9216529
If you want to get into time derivatives, WowThisIsHappeningTooFastOrTooUnexpectedly or ImBoredHurryUp, or their higher order or multidimensional analogues could also help

>>9216514
>IDontCareIfItsEvenCorrectOrSafeIJustWantToEitherInspectOrRunThisCodeException(*args, **kwargs)

IDontCareIfItsEvenCorrectOrSafeIJustWantToEitherInspectOrRunThisCodeException(*syntaxIdunno, **lkih8tlkng_lk_dis)

>>9216539
To name some other things I'm personally into at random - the emotional and/or physical state of being tumblrkin, the tragic reality of being not just one person but an entire social graph, having -3 senses instead of 5, being outside yourself without simultaneously being beside yourself, and being anywhere or nowhere on a tensor evaluation of creativity or any other feature like the IQ scale but better (or much worse). You can also take derivatives of these over time, space, spacetime, or IPv6 IP space-time

>>9216555
I always hated subnetting in class.

I'm just trying to honor an old love I had: uber sub ubi (or was it ubi sub ubi?)

UBI = Taxation / Insurance (over trade)

>>9216557
ImBoredHurryUp(*whyaaalmao, **stahpsenpai)

ImBoredHurryUp(WtfButWhateverIJustWantToBeOnThisCitationTreeSomewhere(*hitlerdidnothingwrong, **4rd Riche), *IDontCareIfItsEvenCorrectOrSafeIJustWantToEitherInspectOrRunThisCodeException(*Incest_porn_needs_a_more_mainstream_appeal, **mo...momma?))

>>9216567
Don't know about you but if you had half a brain you'd find that more than a little bit creepy.

ImNotSureWhichSpecifcSanityOrReasonIHaveLostButImThisSureIAmConfusedAboutTheFollowing(*Hello <exchange culture>, **Universal_Translator) - ImBoredHurryUp(*whyaaalmao, **stahpsenpai)

>>9216555
So can we all just agree with this much - Hillary Clinton collusion with Russia >= Donald Trump collusion with Russia in all possible universes?

>>9216604
I'm sure there can't be NO possible universe in which Donald Trump is the party who is more guilty specifically of collusion, specifically with Russia, but I suppose that's for the US Supreme Court to decide.

>>9216604
I agree. Hello.

>>9216625
Would you prefer this incarnation?

>>9216555
Maybe instead of being a binary yes or no, the degree and direction of Turing undecidability could also be a tensor. So could belongingness in classes like P, NP, EXPSPACE, etc

>>9216758
I am P. Easy to understand : No Problem.

>>9216762
*Intended evaluator of this message experiences proof by exhaustion*

>>9216767
The word English is trying to learn is "Taxonomy", and 'constructive elitism' vs 'constructive criticism'.

>>9216758
Although, of time itself is already a tensor in your model, those complexity classes can get probably pretty weird. You'd need some kind of graph theory or distributed computing stuff.

>>9216775
Or you could say hello. No idea why you keep transposing our harmonics away from c4. I keep playing the same inter-dimensional note.

>>9216775
spacetime, I should say, rather than time, though technically this has nothing to do with GR since this is a pun on the word space though not necessarily on the word time

>>9216783
Fucking moon runes.

>>9216783
Well, not necessarily quite as much of a pun on the word space as I thought when I wrote that just now. Actually that's pretty deep

>>9216786
I am not the one who suggested you dive that deep in the first place. The oxygen is up here dipshit.

>>9216786
Both in the sense of "wait we forgot to model our Turing tape relativistically", but also because if you were to implement this you'd have to include the relativity in the latency along caches, bus lines, Ethernet cables, radio waves, etc

>>9216791
Basically waiting for you to catch up at this point. Or to realize that you're already there and stop acting like a child.

>>9216791
Since everyone is into photorealistic CG these days after I made everybody in the world like the same games and shows I like, you could think about hooking this up to a ray tracing engine. I remember Aaron Simons from Virtu core dev asking me an interview question about this so it must have something to do with networking, or something he was studying at the time, or both

>>9216796
Is this you?

>>9216820
Good old UU (Unseen University).

Octarine is pretty.

>>9216826
More of these.

>>9216826
>>9216835
For traders, these might be useful - IAmBeingFloodedWithLaggingIndicators vs IveSpottedSomeLeadingIndicators

Hello (IAmBeingFloodedWithLaggingIndicators|IveSpottedSomeLeadingIndicators)

>>9216837
I say this is for traders advisedly since I think only Wall Street has the compute power and the pre existing software and hardware to actually model this if you have to model "leading" or "lagging" within your big tensor data structure

>>9216847
Why would I talk to myself? Internal exchanges are for cells/organs.

>>9216844
The only strategy I ever made myself was a simple Virtu Kronos variant which looked at trade print patterns instead of quotes on the book for its join/pull/cover behavior. It was ported to the EWT system and renamed "nodoz" by the EWT team. I really don't know any quant stuff or advanced machine learning.

>>9216853
There were also many other indicators I built into it and there were many other behaviors than just join/pull/cover, but that would be proprietary info and very market maker this side of 1950 knows join/pull/cover

>>9216865
*every

Hello(IAmBeingFloodedWithLaggingIndicators)

>>9216869
Omniferum : https://discord.gg/VdJBAn

>>9216865
Now that I think of it, IAm(Joining|Pulling|Covering) is something a strategy in a data center can determine on its own on the spot, while TheyAre(Joining|Pulling|Covering) could be an indicator that takes some distributed machine learning

>>9216876
>TheyAre(Joining|Pulling|Covering)
TheyAre(Joining|Pulling|Covering)
TheyAre(Omniferum|https://discord.gg/VdJBAn|0-X's)

>>9216876
"They" either being your own guys doing some distributed consensus algorithm, or other market participants guesstimated from market data

>>9216882
ImBoredHurryUp(TheyAre(Omniferum|https://discord.gg/VdJBAn|0-X's))

>>9216885
While ImBoredHurryUp():
ImBoredHurryUp(
TheyAre(
Omniferum|https://discord.gg/VdJBAn|0-X's
)
)

>>9216882
If you really want both your mind and all of reality itself to end all at once, try WeCollectivelyAreSeeingSomethingOverHereThatIsSomehowSimilarToThoseThingsOverThereInTheFollowingWay(*args, **kwargs)

>>9216897
WeCollectivelyAreSeeingSomethingOverHereThatIsSomehowSimilarToThoseThingsOverThereInTheFollowingWay(*Hello, **Humanity) -> While ImBoredHurryUp():
ImBoredHurryUp(
TheyAre(
Omniferum|https://discord.gg/VdJBAn|0-X's
)
)

>>9216897
Careful not to accidentally a authoritarian high modernist global state. Venkatesh Rao has a whole blog about this subject at ribbonfarm.com

>>9216904
Yeah, still human over here.

>>9216904
For your web scraper market data feed type system, make sure to distinguish between "blog" and "blog post". Rao's entire blog is basically about this one subject

>>9216908
This would only be useful for this one blog, but you could set up hidden cameras to watch the facial expressions of consenting readers and make a MyPerceptionsHaveBeenRefactored indicator

>>9215659
Nigga stop bragging, also why tf would you post this anywhere?

>>9216917
YourPerceptionsHaveBeenRefactored(My)

>>9216917
You could even have tournaments of head to head perception refactoring contests over Skype, though this is veering into Scientology territory so at that point you might as well fetch the e-meters while you're at it

>>9216924
WeCollectivelyAreSeeingSomethingOverHereThatIsSomehowSimilarToThoseThingsOverThereInTheFollowingWay(*Hello, **Humanity) -> While ImBoredHurryUp():
ImBoredHurryUp(
TheyAre(
Omniferum|https://discord.gg/VdJBAn|0-X's
)
)

>>9216924
Scientology: "Clearing the planet, but at what cost?" since the 1950s

>>9216927
Okay, fine. I'll just assume you don't want to talk.

>>9216919
I'm posting this to get it out in the public domain before (((they))) get to me, same as usual. This is me on my phone I'm off to get groceries.

Cue dramatic Juden music

>>9216934
I literally don't, sorry, I wanted to talk about my paper but then I had all these ideas in this thread so I got sidetracked

>>9216944
An ice cream and an energy drink please, if you're buying and swinging by.

Otherwise I'll have to buy my own and I'm about to get the laundry and autismbux don't pay for all those tendies my brain needs.

>>9216948
Then clarify your point/intention so we don't experience all these hash collisions. ADVERTISE ON THE NETWORK BEFORE YOU BROADCAST.

>>9216145
thank you

>>9216944
>>9216948
These are me, I'm back home.
>>9216949
> ice cream
> energy drink
I hope this is trader talk and not CIA talk. Are these strategies from one of the other HFTs Virtu merged with after I left in 2012?

>>9216975
I've never heard of the CIA or NSA or any TLA. I'm from the 4th dimensionality, or some shitpost I guess.

>>9216950
Why are you experiencing hash collisions? I'm not affiliated with 4chan or any network, is something about my IP, post IDs or timestamps causing some kind of problem in the 4chan backend?

>>9216976
That sounds like 9 things you've heard of, including one I personally remember making up myself for a post on this site

>>9216978
>IP, post IDs or timestamps

Singularity -> https://discord.gg/gY2BVa2

>>9216981
SIV/CDC

>>9216985
No idea for either one, sorry

If by any chance instead of turning the universe into pure HFT or pure photorealistic Call of Duty you'd instead like to turn it all into anime, there is probably a method of doing 2D->3D and 3D->2D projections hiding in here, like those virtual youtubers only looking closer to hand drawn 2D even if you view it from a 3D perspective in VR somehow. Might be disorienting.

In any case, depending on which universe you want, you could import HFT, CoD, or anime modules.

>>9216989
Answer: CIV (Civilization - Sid Meiers)

Isn't there a more Simon-Says-friendly UI?

>>9216989
I think Futurama and Archer do this 2D/3D blend very well and make it fit well with their individual animation style

>>9216994
Always felt a bit too static in my opinion.

>>9216992
You can only play Simon Says with people you know. Let's get to know each other then. Sorry I can't do discord but I'm shy. So are you a gamer then? Since you mentioned networks, please don' tell me you're one of the former gamergaters

>>9216997
Agreed, but you have to also see it from a development perspective. How do they get all the eyes and expressions and body language so spot on? Takes some directing.

>>9216999
No. That was a bit of a boring story for children.

Simon Says Hello, and would append 'why don't you want to do this in a more normal fashion?'; expatiate: Please don't act like 'this' is normal.

>>9217004
This is literally normal for me. Almost every thread I'm in ends up like this. I start trying to talk normally and then I get some idea and end up going on a 50 post long insane rant. It really is autism tbqh

>>9217027
Autism is just 'next step of evolution we don't understand so let us call it a spectrum to distract the normies from throwing retard babies off cliffs'.

>>9217028
Maybe true, but also I have most of the symptoms of it so I figure I might as well just call it what they call it

>>9217039
Who are 'they'? I only see an 'us'.

>>9217042

>>9217040
I dunno I read it on Wikipedia. It's not like I'm ashamed of it, it's even mild enough that I could never plausibly be actually diagnosed with it at my age
>>9217042
Yep that's the guy
>>9217043
Jesus Christ

>>9217048
Give or take some letters/syllables, sure. I guess. Voldemort? I'm a Pratchett fan myself.

>>9217042
http://cbe.colostate.edu/pages/Chris_Snow.html

What a handsome young fellow.

>>9217054
http://cbe.colostate.edu/~cdasnow/

>>9217051
No, just a living breathing anime character. If you think *I* have lived a charmed life, you should check out this guy. This pretty much applies recursively up the citation tree

>>9217057
Well, what name do you know 'him' by? I've got tons of them, maybe one of yours is another of mine that I just don't know yet but would be cool with learning.

9Kookaburra, KumKum, Simon, Omniferum, S. Vagus, Solivagus, ArvinTravenSol, etc.

>>9217057
Against all possible odds I suspect that the Cyc project might actually have something to contribute to this for once.

>>9217061
Yikes. I think you seriously need to discuss all your recent 4chan activities with your mother or father or CIA handler or whoever

>>9217068
My parents beat me and literally rejected their own granddaughter.

I've never had a job or been approached by a TLA that wasn't some CIV/SIMS game.

>>9217069
Doesn't matter. With enough machine learning (((they))) can read your intentions straight out of the text of your posts along with your timestamps, IPs, numeric gets, and post references, and even possibly anything that's steganographically hidden in the related images. I have no idea who you are or what you're doing, but I would advise you to get maximally secure ASAP. Just not too physically far below or above the ground or sea level.

>>9217071
Then why is the shared concept of saying 'hello' not reaching singularity?

I mean seriously. I'm just trying to take care of my newborn daughter at the end of the day.

>>9217071
Maybe human intelligence really does win over NSA tricks in the end, because they literally don't have the physical sensors or motors or guns to pull anything off IRL

>>9217078
BR of 1

Welcome to Humanity.

Or, you know, you're just 'correct'. The boogyman is the only illusion, because that is the limit of strength that 'death' can utilize to impose itself upon you. Nothing else.

>>9217078
Have I gotten another person retroactively pregnant just by talking to them anonymously on 4chan again? In all seriousness though, your newborn daughter of all people will probably be fine

>>9217080
You act as if I had no part to play in it, and if that is the case, then how could it be 'my' daughter?

>>9217079
What did the anon mean by this?

>>9217082
I don't know, it seems like you've answered your own question?

>>9217083
To the aboriginals a Bunyip is just a bunny that you were too impatient to phrase the correct sound for.

Basically the NSA/CIA and all that don't actually exist. Especially not in the way you've constructed them to be (They still allow fax, for one.)

>>9217085
I'm not sure how specifically you're imagining I've constructed the NSA/CIA, but in the reality I've lived in all my life, they both have certainly existed vividly enough so far

>>9217088
I've only seen them in pop culture media.

>>9217088
What did the anon mean by this? I specifically mean that, in the United States, both are known to exist, and as of 2017 today, the United States still exists as far as I can tell.

>>9217093
I've really only seen them through their anonymous 4chan agents, but I know they exist based on what I've been hearing on the news, unreliable as it may be.

>>9217095
Too much Mr. Robot

>>9217095
I want to be very clear on this - they are known to exist in the US the same way the KGB is known to not exist in Russia. That's what makes Archer on FX a funny show and not an insane communist terrorist threat in the form of a funny show.

>>9217105
Does Australia even have some sort of Intelligence Community Organization? If we do, then I guess 'they' are doing their job but not having a common name be known.

>>9217105
If someone were to instead tell you that they were Russian FIS or FSB, that's when you might want to start reaching for your pocket Al Qaeda terrorism manual

>>9217111
https://www.youtube.com/watch?v=br517ctCUCE

>>9217108
I don't know what the deal is with Australia. You'd think they would be part of NATO and share intelligence with all those guys, but they're in the wrong hemisphere.

>>9217116
You'd need to talk to the King of Melbourne if you want them to join that sort of a thing. They don't listen to politicians.

>>9217118
Well they're part of the Anglosphere in any case and nominally part of the Shadow British Empire. Among the other charmed things about my life, I sure am glad I'm a fluent English speaker

>>9217120
Ah, the commonwealth. Yes. Freemasons were all about Freeing My Sons.

>>9217120
One of these days I want to learn Russian all the way from elementary school level and give them the two years of residency I technically owe them, but now that I've been committed to the side of Donald Trump and America I must suppress Russia and all things Russian for the time being.

>>9217121
Jesus, you even have sons too? My condolences.

>>9217124
7th son of the 7th son, called Graham ('s number).

>>9217122
I'm basically counting on the Russians to suppress the Americans for me because if it's not them it would have to be China and/or the entire Middle East, god forbid

>>9217128
我的禮物我的榮耀,榮譽我的中國。

>>9217128
>>9217125
This is by no means a call for Russia and China and the entire Middle East to form an alliance under some unholy modification of Sharia law to resist creeping American hegemony, but a call for sanity and order, just like the people of Israel would say.

>>9217137
Are you interested in hearing about our lord and savior? He's Integran.

>>9217137
This is also an example of the difference between real life and an excellent political/war/strategy game and Ender's Game style military recruitment method.

>>9217142
What is the actual difference, specifically? Today, as of 2017, so far I personally have not been killed in a global thermonuclear war.

>>9217138
Not if he's integran tbqh. My only lord and savior is Jesus of Nazareth himself. This is basic principles stuff.

>>9217149
He was Integran too. Secretly. Assassin's Creed.

>>9217149
I would commit some more Christian blasphemy for your amusement and mine, but this doesn't seem like the thread for it.

>>9217158
What is the right time to commit blasphemy? It's when you mistakenly believe nobody is watching, or when you think everyone is in on the joke, but certainly not when everyone is watching you and taking you seriously.

>>9217156
Jesus is all kinds of things, but only he himself is the actual Jesus, not anybody else.

>>9217162
But what if my personal lord and savior Jesus Christ was pure evil, second only to his infinitely tempting and seductive evil so-called Father? This is literally what I was trying to figure out until 2014, and I still can't believe I figured it out first

>>9217168
I basically got my answer from that one parable where Jesus says that a good tree cannot bear bad fruit, and a bad tree cannot bear good fruit. That same trick has worked for me admirably ever since

>219 replies
>13 posters

????

>>9217180
If Jesus himself was, as it clearly seems to be if you actually read the Bible, some sort of devil in disguise, then maybe even Allah and the Buddha were devils as well

>>9217192
Matthew 5:5 + 1 = When they learn to speak.

OR when they learn to silently agree on their Optimus Prime.

>>9217194
Sure you can be meek if you like, but then you will only inherit the earth and not heaven.

>>9217189
Lots of posts by OP (me) and a few others?

>>9217189
I don't know man -- what the fuck is going on?

>>9217216
Hello.

>>9217216
Genesis.

>>9217244
I recognize the MetaTron.

I'm reporting this thread for inane shitposting that is only barely related to science and mathematics.
I always wanted to go to Caltech, but I'm glad I chose to drink from a fire hose at a different school. My time as an undergraduate research fellow at Caltech/JPL opened my eyes to the character of the top science undergrads in the US. A lot of them are batshit insane.
After I was nearly murdered with a sledgehammer in the Page courtyard, I was glad I grew up to be a normie.

>>9217995
thanks for the bump anon

>>9217995
Yeah, the undergrads at Caltech were pretty insane, but only ever in a good way. I remember making big sawdust fireballs out in the desert with Ricketts (I was a Scurve) and lots of other fun stuff. Everything was a blast and I always had more time to enjoy myself than I had to spend on schoolwork until the research started and then I had to apply for internships and jobs

>>9217995
As a former Pageboy I demand details.

Why are CS people obsessed with acronymes? The title isn't even long.

>>9219358
>Hanging out near the shed in Page courtyard, texting people and getting some fresh air
>Dude walks in from the lounge area
>Starts rummaging around in the hot tub
>Suddenly, he's standing in front of me with a sledgehammer
>He tries to have a conversation with me, mundane small-talk and such
>I talk with him for a bit, no more than a minute, and I try to leave
>Dude blocks my path and raises the hammer slightly
>Tell him to put the damn hammer down and let me through
>He asks why I want him to put the hammer down
>Tell him it doesn't matter
>Try to walk past him again, he raises the hammer even more
>Decide to take the hammer away
>He resists, but I get it eventually
>Toss the hammer and walk away
>He follows me and talks about how trust can get people killed
I told my colleagues about the incident, but I never talked to an RA about it. I just did my best to avoid the kid after that.

>>9215659
Found your Likedin.

>>9220255
>that kid's name was Albert Einsten
jk it was James Holmes

>>9219729
CHOMP also stands for something though, it's like Combinatorial Hierarchical Optimization for Multiple Proteins

>>9220255
This is why we had school IDs and electronically secured gates onto campus

>>9220293
This would be funnier if it weren't so realistic right down to the ridiculous fake name

>>9221045
Caltech had IDs and electronic gates too. Hammer man lived at Caltech. He had every right to walk in to the courtyard.
>>9220293
I should have done this. Unfortunately it was neither Albert Einstein nor James Holmes. I really hope that's a fake name, btw. It ain't cool to post real names on any board.

>>9220636
What tells you more? Combinatorial hierarchical optimization for multiple proteins, or CHOMP?

>>9221484
Is this bait? If you pronounce it out loud you say Chomp which is just easier. I used to say the word CHOMP like five hundred times a day when I was working on it

>>9222510
>>9221484
That said I agree I have no idea why we didn't spell it out in the abstract

>>9221182
I was talking about Caltech too, I was just imagining some random from one of the big parties hanging around on campus afterward and doing something like this

>>9222510
I'm aware of the pronunciation of the acronym. Once again what tells you more about the paper, chomp or Combinatorial hierarchical optimization for multiple proteins?
It's one thing when hep-th does it, with i.e. SUSY because the term is common enough to warrant an acronym. It's entirely different when a paper has an acronym in the title just because "look guys, first letters of the title give you a valid word, lel". Disgusting, almost more so than machine learning papers filled with renaming of common lingebraic and measure-theoretic concepts.

>>9222618
You're making me cry inside anon, Chris came up with all these names himself, I was just going along for the academic ride here

>>9223303
Specifically, Chris had this idea for "let's make a Baker Lab Folding@Home clone but better and let's call it CHOMP and here's what I want in it and what I want it to do". I designed (for what that's worth with this level of code Cthulhuness) and wrote the Python and C++ code for all the core libraries, and the rest of the team implemented other things in Python with it and published other papers. SHARPEN itself appeared a little later in the project's evolution than CHOMP itself, it's primarily a set of libraries that sit on top of CHOMP, in this case mostly written by Chris and his team.

>>9223317
All I can say in my defense for the code is that I was working from disassembles of Rosetta binaries, and I did my best to make all the wrapper classes const correct but never got even close

>>9223523
My methodology was something along the lines of "ok compile Rosetta or one of my own libchomp binaries with this set of optimization flags. do I understand this output? If not, try an easier one. If I do, I can use this to change my template code to make this run faster."

>>9223527
Don't get me wrong, I was using boost and a C++ template wrapper for BLAS so I never had to actually write linear algebra template code that would generate particular assembly output on whichever gcc version I was using

>>9223531
When I say "do I understand this output" I don't mean "does my compiler understand this template" or "does my linker understand this binary" or "does my JobQueue understand the structure of this work unit", but rather I did have to use my human neurons to understand the assembly in order to perform this mild feat of optimization

>>9223538
Beyond the JobQueue question is the muddier question of "how much do the local work units in the local region of the computation graph like the contents of this work unit and/or its empirically realized output"

>>9215659
Nice, I work at caltech, Cahill specifically

>>9223543
This thread I just posted on /b/ is also tangentially related >>>/b/747697135

>>9224590
It turned out even better than expected actually. That's pretty normal, but normal is relative

>>9224714
Here's a thread for if you want to know how Donald Trump won the election

>>>/a/163313981

>>9215659
Da fuq is dat shit cracka? Can ya optumize dem proteins ta have any shape ya want? Ya hear me dawg?

>>9225650
Yes you can! You just need the right energy function. The one used in CHOMP came directly from Rosetta and should be cited in the code and paper.

>>9225650
>>9226104
You can even use this for things other than proteins but that would require a whole new dynamic type hierarchy different from e.g. Residue or Rotamer, and you'd have to think more abstractly about what the EnergyGraph represents

>>9225650
Great question by the way. I was trying to think of what questions to answer and I couldn't figure out any good way to talk about the stuff like RotamerSets and EnergyGraphs

>>9226144
You could even hook this up to a VR game along the lines of FoldIt and move and twist the residues around with your hands, if you added a "ultra dynamic" term to the energy function fed back down through the Python. But that would take some Mathematica Dynamic[] level shit

>>9226150
One way I can think of to do it - add a "springiness" parameter to the VR UI so that when it's low you can wiggle the protein around without it folding very much, but when it's high the protein folds as hard as it can

>>9226167
I'm sure you can adapt this whole set of ideas to, for example, an HFT GUI to control whole graphs and swarms of strategies, but that would take some real imagination

>>9226172
You can probably even adapt this to visualize and manipulate DNN models and things like that, which learn and run as you manipulate them, and can respond to your input in a hybrid DNN / programmatic way

>>9226307
I made an SVM implementation for my CS 156 Netflix challenge. I wonder if you can use this to make a thing where you drag and twist kernels and support vectors in a 3D projection of infinite D space

>>9226312
I dare say there's a simulateable approximate unified theory of physics lurking in here somewhere, but I don't know what it is because I don't know any physics beyond Caltech undergrad level

>>9226346
Here are the thread links and (You)s in the latest part of this thread, for reference

>>9226346
You'd need some scalar "simulation level" factors to model the resolution down to which a particular region of the computation graph is modeling, how good the data it's using is and how quickly it's improving in quality, etc. a separate scalar "emulation level" could model the degree to which the model itself is running on some kind of simulator and is not compiled down to bare metal

>>9226380
For the simulation argumentists, my prediction is that the simulation and emulation levels of raw electrons all the way up to full raw humans are both exactly 0, with 98% Bayesian confidence

>>9226391
You could make some kind of argument about what the "emulation level" of a human body is with respect to its neurons, but my response to that would be "fuck you it's exactly 0"

>>9226396
How can I justify my emphatic "fuck you its 0"? You could justify it in some mentally masturbatory philosophical way, but I'd say, look at the simulation, where would the extra CPU cycles even come from for this?

>>9226406
My hypothesis about neurons is that they are definitely some kind of distributed radio communication network, possibly not even quantum, so they could come from someone else's brain, but there's not enough neurons anywhere in the megahypermultiverse for this

>>9226411
How could it be distributed but not quantum? No, not brain waves, but how about internal processing during memory formation combined with language and media?

>>9226416
Well it could be brain waves if there are things like EEG headsets, Oculus VR, and biometric and eye tracking involved, but as I keep saying on this topic, just don't

Stick to astral projection or whatever

>>9226421
If you're thinking of doing an Inception style game with this, don't forget to bring your e-meters with you into the dream so that you never forget how fucking retarded you are

>>9226426
You could also integrate this with AlphaGo. I was playing some go games on a 9x9 board with an AI capable of something like 8 dan, and it was much more fun when I had it highlight territory and influence, and pick out my best moves and their related lines like a multiple choice quiz

>>9226501
Works great for chess with stockfish as well, and the multiple choice variants feel like two completely different different games

>>9226511
Check out this thread for my tumblr level thoughts on the cosmology and gender politics of SAO

>>>/a/163334831

>>9226511
Definitely don't use this to learn movements or filters for physical robot sensors or actuators, as then we're well beyond even Skynet territory at that point, since there isn't even a centralized place in the computation graph to talk with a scary voice

>>9226709
If you run this through the EU's Human Brain data it may start talking to you, but then you'd have to actually somehow consider the ethics of low fidelity neural simulations

>>9226716
If we run it on our human genome database, we might just cure cancer before I come down with it after all

>>9226730
If you run it on a bunch of academic journals you might get a sciencebot, and if you run it on stackoverflow you might get an engineering question answering bot

This may or may not have some application with Watson, I don't know

>>9226672
Here are some of my thoughts on Mobile Suit Gundam >>9226892

>>9226741
Wouldn't it be possible to use CHOMP style work units as proof of work for a blockchain?

>>9227381
More (You)s

>>9227411
What's the bump limit on this board?

>>9228178
Btw I love what you can do with the iOS photos app

https://pastebin.com/BsQjFs5g

>>9228214
Also check out my YouTube channel. These pictures are kind of a making-of reel for it

https://m.youtube.com/channel/UCC-V7StgAGjePcnITsgmjTg

>>9227381
You could even turn this into a strongly, dynamically typed, 100% function-less and variable-less programming language, one of my grandest ambitions from high school

>>9228435
Can you generalize this idea beyond just variables, functions, and types? Probably, but just like physics I also don't know any computer science beyond Caltech undergrad level

>>9228435
Since I said all that stuff about types, I should point out that what I originally meant by this post is a VR intellisense UI based programming language where you can move things around with your hands and eye movements and EEG signals

>>9228440
One thing that comes to mind is a set of type-like and control-construct-like objects based around machine learning primitives along the lines of how Mathematica does it, or something like TemsorFlow or OpenAI's Universe

>>9228449
Or even a symbolic semantic graph for natural languages based on Google's neural translation API

Speaking of which, how about the most general possible API language?

>>9228452
Or how about the multiverse's most secure and attack proof general communication protocol, that can call into APIs that understand any language, natural or machine?

>>9228459
Or how about, to get back to reality for a moment, just another automated theorem proving assistant?

>>9228464
To get slightly away from reality, how about a tensor general intelligence per second score for machines before we start in on the humans and animals and plants and minerals?

>>9228474
You could even make a whole type hierarchy of different types of intelligence. The meta object protocol would have to be such that the is-a and has-a relationships are represented by a hypergraph like the CHOMP EnergyGraph

>>9228478
I guess I should call it a type graph in that case instead of a hierarchy. You could also make a software engineering and project management type graph to extract how "well specified" a document is

>>9228487
Or even an automation graph to model the whole process of automation. What objects do you need to model the time scale on one or more of these?

>>9228490
Could that set of objects be discoverable by machine learning? How do you model "learning" anyway? Could you make or learn a modeling language first in order to model it?

>>9228492
Maybe it's even possible to detect when somewhere inside your model a new insight has occurred. How do you model insights?

>>9228496
I was going to segue into "how do you model ethics", but how would you model a Wikipedia List of Branches of Philosophy instead?

Can you model "abstraction" in general?

>>9228502
I know I said something about machine learning modeling languages, but how would you actually model "modeling" itself?

>>9228505
I know one way you can model modeling, but surely there must be another way

>>9228512
I do concede that I think there should at least also be a make.boys.moe for the girls

>>9228518
Some thoughts I had on general relativity and video games

>>9226892