/sci/ - Science & Math

>>10217742
how can one be too stupid for engineering

I'm not CS but "hate threads" are just underage autism

>>10217751
>t. CS

>>10217746

ITT: People with comfy jobs thread

>>10217753
You have to be 18 to post here.

t. licensed mechanical engineer

>>10217742
Fuck off faggot. The only reason you make these threads is to cope because STEM's career prospects fucking suck. I deeply regret studying engineering.

>>10217753
No, just got tired of edgy kids trying to compete to see who can hate everything the most at the start of the decade

funny thing is that most of the math/physics undergrads here who think they are special, will get into academia, become professors working on bleeding edge work on CFT/string theory/whatever will end up having to jump ship at some point and work for a tech company as a data scientist/programmer

>>10217742
>CS
not science or math

My uni had a theoretical CS Masters program, I actually thought of it but they shut it down too soon for me to apply, is theoretical CS good-tier?

>>10217790
that doesn’t make CS and Eng people not brainlets, brainlet

>>10217917
A masters in theory? I'd be incredibly wary of that. Any good theoretical CS program is a PhD program. On that note, a good CS theory program is a top tier math/science program, period, since it's on the cutting edge for both pure theory within its own foundations. for theory research across disciplines, or for application. Susskind, Aaronson, Erickson, etc. are prominent theorists. I personally like Complexity theory and how it relates to analysis: https://arxiv.org/pdf/1305.0453.pdf

CS is a supercategory of maths

>>10217742
Are we talking undergrad or grad? I could believe the majority of undergrad CS students want to be codemonkeys, but grad CS is legitimately boatloads of fun. If you haven't figured out that theoretical CS is how math departments reap more funding from various organizations, then I don't know what to tell you

I'm going to study AI, data science, blockchain, operations research, and literally any other meme I can get my hands on while I major in either physics or applied math. Change my mind.

>>10217751
It's mostly started by engineering majors who don't want their employability usurped by the hot new trend. If not their employability, then the 'rigor of their program.' It's a rat race to make themselves feel better about what they're doing. Math and physics majors jump in because it's easy to make fun of people if you paint them all like brainlets.

Engineering majors tend to do these comparisons based on the number of required classes and their apparent difficulty. I've looked at a lot of good CS programs outside my old undergrad, and it seems that the difficult classes are generally electives. In general, I feel this is true for any STEM, non-engineering major: the rigor of your degree really depends on how hard you want to go. I do think the majority of CS majors do the bare minimum to codemonkey it out in the industry, so there should be a separate major for them to do that. However, the people who stick out and take the hard classes (and often time the grad level classes) are by far some of the most brilliant people I've met. Few and far between. So while I'd say that the average engineering slacker is smarter than the average CS slacker, the CS boi who goes ham in his classes, takes grad classes, double majors in math or physics. etc etc. outclasses the majority of engineering students, who are really just looking for a stem oriented degree in bachelors so they can make money. But hey, I think that really means that you shouldn't base people off their major and instead examine what classes they chose to take and what research they did.

\blogpost

>>10217947
I mean, these are really applied subjects. I'd suggest focusing on math or physics. Those are pretty key to doing well in any CS topic worth the time. What you listed is mostly software engineering, self made code skills, some basic optimization theory, etc.

>>10217746
good question. i know double digit iq retards that got an EE degree

>>10217947
/g/'s meme > AI, data science, blockchain, operations research, and literally any other meme. And I'm not even kidding.

Be more heavy in pure math since it teaches better ways to deal with abstractions.

If I would start over, I would first learn pure mathematics and a multi-paradigm programming language like C++ or Lisp,

You should learn the matured knowledge first before going blindly to the latest fads. Most of the latest fads in CS will have an inevitable decay.

>>10217957
Thank you, you virtuous person. You have properly explained the cause of their envy, which ultimately leads them to arrgogance, and then hell. I shall copy this, and spam it every time i see hate threads.

>>10217957
If they were smart, they'd undermine CS careers by helping push for every K-12 student to learn actual programming skills.

feels good having job security for the next 20 years
maybe one day you'll find a job OP :^)

Is it truth EE majors can easily get CS jobs?

>>10217917
My university had an accelerated master's program for CS. It was basically just the undergrad course load, but with graduate level versions of courses. It was a highly competitive, theory + math heavy, program, and pretty much anyone capable of doing it made ~15-20 % more than those who simply did the undergraduate CS upon graduation. It took me one additional semester to do, and I think it was worthwhile.

>>10218101
Companies often hire general "smart guys." If you can code and have a degree in engineering it's not unusual to get hired as a codemonkey. It's not as easy as someone with a CS degree, or, even easier, a person with a portfolio and a track record.

>>10218067
Programming and CS are two different things. Programming jobs, I agree with. CS jobs are either in academia, national labs, or at the R&D departments at companies like google or lockheed martin. I'd say HPC and cryptography get passes for "CS jobs" as well.

>>10218139
you don't even need a college degree to get hired as a codemonkey.

half of the programmers at my company never went to college

>>10218030
Basically. Learn hard theory and hard systems, as they hail from math and engineering respectively, which are the foundations of many subfields of CS. With these two, you can tackle any other problems.

>hello sirs I am the most qulified coder for job.
*hands you mill degree from dehli*
>I can make loops and automize program
*shows you his snake game code made in python with no comments*
>Thank you sirs, I can start moonday
*gets hired because the manager is also a filthy pajeet*

>>10218162
lol

>>10217742
Hey fuckers, I learned Djikstra's algorithm today.
Bet you're not as smart as me.

>>10218139
You are talking out of your ass. I have literally tried this and it is not true.

>>10218261
Stop pretending to be an EE, csfag

>>10218269
I am >>10217779 you little shit.
>inb4 ME
I'm a controls engineer and do a lot of programming.

Interesting fact: in India, the Calcutta institute for tech designed a computer made entirely of human feces!

i will program slave bots to take your job and another one to take your wife and look after your kids. fuck ill even make the bot that sells you the gun that ull end yourself with. u worthless faggot hows it feel looking at gay stars and other mens balls. meanwhile i got one hand on a bitches thigh and the other on my keyboard replying to the 5th university begging me to come study there to the point where theyre willing to pay me to come

/sci/ is cancer

If you your concern is getting a job, then become a nurse.

CS grad student here

we're working on technology that replaces engineers. in particular, we're working on AI that designs circuits and filters for signal processing.

It amazes me how big one's ego can get that would lead them to make a thread like this, and have its discussers take the matter seriously enough to think that it has any relevance to anything important at all.

Remember, science is just the application of rules to beget more rules. It is nothing that AI won't be able to do. So don't think that just because there are people that struggle with stuff like this that you're suddenly on high ground because artificial intelligence can do it, too.

>>10218328
I'm a CS grad in theory. I know there is research is automated assistants into circuit design due to circuit complexity being a famously well researched field, but there's a good amount of research that just throwing a feedfoward neural network at won't solve.

ML/AI is not a bad field inherently, but it's so easy to do dogshit research when you have so many people who don't give a shit about what the models are actually saying. We need more AI theorists and fewer "I wanna throw AI at this issue" grant chasers

>>10218364
AI is not the end all-be all of CS. I don't want to get into the AI debate since I doubt anybody here is an expert and can actually debate past the basic buzzwords, meme infographs, and "my favorite industry tycoon said X about AI," but the short answer is that automating repetitive tasks is easy, but automating a list of responsibilities isn't so easy, and most numerical/ML techniques can be finicky if you're not careful with narrowing the problem space down

>>10217742
He says, not making a better app than omnichan...
#BringCloverBack

>>10218435
Making apps is programming and >>>/g/, not CS

>>10217742
>posted from a COMPUTER
/thread

98% of all CS majors are just codemonkeys that don't need almost anything from their own degrees. There is some depth to the field but then you are going outside of CS anyway into something like Math.

ML is the most braindead and misapplied way to "gain knowledge" that is also extremely easy to do. So you make these extremely specific models and just train train train to get a conclusion for a grant application but the research itself can't be generalized anywhere except your very specific problem.

>>10218489
>but then you are going outside of CS anyway into something like Math.

No anon, that something was CS all along. The difference between math and theoretical CS is one of motivation and grant proposals. Not just from the foundations, but just research for research's sake. Literally the only difference is usually rephrasing the question and giving some motivation to the problem through existing theory

>>10218483
>t. stereotype CS major taking credit for EE and CE

Is there even a reason to do CS over math in undergrad if you want to do TCS? Getting a BS in math and then doing CS in grad school sounds way better.

Undergrad CS is full of code monkey courses like databases and software engineering, the only courses worth something might be theory of computation and algorithm analysis but they don't go that deep anyway so they are pretty trivial to self learn if you're a math major.

>>10218523
>Undergrad CS is full of code monkey courses
only at shitty schools

the best thing to do is double major in CS/math

>>10218523
As far as algorithms go, the rabbit hole goes deep and it's nontrivial. What they teach you in an introductory algorithms course is very straightforward, but everything else is full of hard tricks/deep insight. Any intermediate algorithms course that covers harder material like using a fast fourier transform to solve pattern matching or anything related to cache obliviousness makes this apparent.

Theory of computation is a vague word for a large field. It has to do with foundations, automata theory, complexity theory, type two effectivity, etc. Honestly, the best course of action is to double major in math and CS. Focus hard in CS theory courses, but also take hard systems courses (not software engineering courses, but things like OS design, embedded systems, processors/VHDL, etc.). I will say that CS isn't all discrete, and a lot of your continuous mathematics like metric spaces and analysis come back, just with different names like valuation and Levenshtein distance in domain theory.

>>10218523
>>10218552
Also, I dunno where this software engineering course meme came from. I have literally one industry focused class in my entire curriculum. The rest is either requirements in the math department and then specializations in theory and systems

Do CS people study control theory at all?

I guess they don't since they don't have to take differential equations?

>>10218704
Those who want to do computational robotics and mechatronics generally do. At the grad level, it's not uncommon. In general, differential equations are the basis of stream calculus (used to analyze data streams), develop ML, solve some applied problems, do work in physics (either computational or theoretical), etc etc.

>>10218704
Had options for it at my uni. We have a very large CS department though so I'm not sure if it's typical. It certainly wasn't required.

>>10218704
EE here I love controls. Sooo cool

>>10218489
honestly so far i can't see any useful applications from machine learning to biology.

correct me if i'm wrong but it appears the utility of computers to the life sciences pretty much ends at RNA-Seq algorithms and clustering trends from large datasets.

>>10219075
Learning on protein folding is pretty huge
Computational medicine is getting big, though they're gonna need strict testing on its results

>>10219075
Also, algorithmic study of animal behavior. There was good research into anthill construction as a distributed computational task. Basically, we could apply all the theorems and systems knowledge we have about distributed networks to study how ants coordinate fairly simple internal construction algorithms to make a sophisticated final structure

>>10219075
One example recently, by feeding in enough organic molecules, the program could estimate the likelihood of new compounds being useful/stable. I think with proteins it could go between DNA encoding, protein folding, and stability.

Basically with whatever you want to call them, given enough information, they boil down the information to the most relevant information and best predictors of goals. With machines that can experimentally verify results, they can improve their modelling and every iteration is more useful.

>>10217957
you would think they'd be shitting on 'software engineering' as 'not really engineering' not cs which has nothing to do with engineering industry

>>10219075
many healthcare devices contain computers

>>10217775
>tfw just signed for $81k as software engineer when I graduate
I love CS

>>10219374
>>>/biz/

>>10219380
>I-I don't care about m-money!!

>>10219488
Literally not science nor math

>>>/biz/

>>10219491
you probably think CS isn't science or math either am I right

>>10219491
Autist