Master’s Degree in Computer Science
evanp.me
evanp.me
UT has this nice list online that they never truly offer - you have at most 3 courses to pick from each semester. And to finish in 2 years you have to take the summer course, which there is one class and no options to pick from. Last summer I took algorithms… probably one of the first CS courses you ever take. I have to take whatever classes are offered every semester to actually finish in 2 years.
The program is disorganized, with you not knowing what classes offered are till a month before registration. And the program scheduling is awful. It’s a Friday and Saturday once a month. It has consistently been on every 3-day weekend.
The program does nothing to accommodate full-time workers despite its description. The courses are re-packaged regular MS courses, complete with all the homework and office hours in the middle of the work day. Sometimes you’ll have projects that are awkwardly due, due to requiring information from lectures once a month. And you will spend time doing tedious projects - most of them are not well-designed to reinforce the course material and end up being busy work.
All in all, it’s not worth it. It has marginal value a few years out if you don’t have a CS undergrad, but it’s value drops quickly with every year of experience.
I will say certain instructors like Vijay Garg make the program feel worth it. You learn new material, read research papers, and have implementations around research papers. His courses were the reason I stuck around. I took his advanced algorithms and distributed systems courses which were great.
Can you describe how you were able to do that?
1. Courses -There are a number of courses now where you can create projects and publish papers. Again, not the default, but if you put in the effort it's doable. To name a few, Computing for Good, Big Data for Healthcare, Deep Learning, Educational Technology, Human-Computer Interaction and others.
2. VIPs! These I consider one of the "hidden" gems of OMSCS and what I personally did. You can see them here: https://www.vip.gatech.edu/ but basically instead of a class for a semester (but counts as a class for credit and graduation purposes) you work for a professor on a project with several other grads/undergrads. I did one this past semester and was asked to stay on as a research assistant over the summer doing work in NLP. Really excited for it actually and has been a fantastic opportunity. I put a lot more details here - https://redd.it/u6cj5z
3. Master's Project/Thesis - This is the option to do a Master's Project or thesis instead of just ten courses, but it does require more work from the student to find a professor and do that. There's many more details here https://redd.it/9t48b2 and frankly I wish I had done so, but too late now.
I have a Bachelor's degree from a decent state school in California.
At the time, I felt the courses were too easy, and my four years were a bit of a waste of time.
I remember going to a graduation party, and you would think we graduated from med school. The graduates were carrying on like it was so rigrigrous. (I have a very, very average intellect too. I actually flunked kindergarten.)
To this day I will not denigrate the school in any way.
Am pretty sure there’s an interesting story there somewhere :)
Children as young as four in the province are evaluated by educators on (pages 306–308): the development of the ability to interpret and respond to basic communication, demonstrate independence and "self-motivation" in learning, giving and accepting constructive criticism, developing problem solving skills ("e.g. trial and error, checking and guessing, cross-checking), personal hygiene, self-control of emotions, assertiveness when feeling safe or uncomfortable, and other skills. Indicators that educators look for include phrases from children such as "I'm really frustrated" (page 161) as a demonstration as an awareness and ability to label emotions; "I put my vehicle on the shelf so it would be safe" (page 167) as evidence of problem-solving ability; and persistence in difficult games (e.g. card games and outdoor children's games).
I was also curious whether education in kindergarten could actually have a causal effect on improving long-term outcomes. I couldn't find an immediate conclusion on whether or not interventions are effective, but in at least one paper, it's treated as established knowledge that interventions work. From a quick search, a longitudinal study published in a paper called "Task-Oriented Kindergarten Behavior Pays Off in Later Childhood" [1] with 2837 participants showed a correlation between self-regulation skills in kindergarten and long-term outcomes. The researchers wrote that "early screening by teachers [in kindergarten] introduces the possibility of preventing future learning and behavioral difficulties." They also asserted that "classroom engagement is malleable and amenable to interventions."
While causality was not clearly established in this paper alone, it looks like a reasonable prediction from the correlation that improving classroom engagement as early as in kindergarten could plausibly lead to better life outcomes in years later in life.
TL;DR: While kindergarten in Ontario in recent years may have different expectations than the commenter's time in kindergarten, it appears that educators do evaluate young children on general life skills (e.g. self-control and ability to be aware and label emotions) as a screening tool, and also potentially for interventions to improve engagement in the classroom (which could plausibly lead to better outcomes in years later in life).
[0] PDF, 2016: https://files.ontario.ca/books/edu_the_kindergarten_program_...
I remember all the children had to go up to the calendar on the blackboard once a month and put the day, date, and year The teacher would whisper in our ears where to put the plackard on the black board. If you failed--you had to do it the next day.
I just couldn't remember what she whispered in my ear while walking up to the calendar on the chalk board. Looking back it was nerves. We just had to move the pre printed plackard to the right spot.
The children used to yell "right--left. I would go to 30-31 places until they clapped. (The kids were on my side. They wanted me to succeed.
I was frozen with embarrassment though.
I think I had some learning disability, or emotional problem.
I just remember I had a hard time learning. In all honestly, I just wanted hide in the playhouse from the other children. I was basically very shy, and nervious.
That's when they put me in Early Primary.
It didn't bother me because 1/3 of the class was with me.
My family moved two years later to San Anselmo.
I was happy. My father, and mother were happy. My dad bought a four bedroom home. We all had our bedrooms. I love life, but my family more.
When I entered 2nd grade, I became worried. These new kids had no problem answering questions the teacher would ask.
I------would just cower in fear hoping she would ask me anything. Well--she didn't overlook me, and I would just freeze in fear when she would ask me to repeat what she just said. (Looking back it was basically nerves, but maybe a learning problem? 99.99% of me now feels it was just nerves.
Ok--it's the 70's, and teacher, and my mom, had an unusually long conference.
They held me back again. They put me in this worthless Speech class. I was a studder. "A, E, I, O, U, any Y." Repeated in a route manner. It made no sence to my young self.
I did have a younger sister whom would remind Everone that we were three years apart in age, but 1 year apart in school. There was always a silence from intelligent adults, or an uncomfortable silence from the rest.
By the time I got to the 3rd grade, I knew I just could be held back again.
By the time I was in 7th grade, I was a B to A student. Everything just clicked in.
High school seemed easy, but it was only until my last year I took it serious.
Most blue collar kids knew nothing about good/bad colleges. My sister, and myself, knew nothing of the SAT. It probally didn't help that my Electrican father thought "College boys were tax dodgers." (It was a different time in America. If you didn't go to college, it didn't matter. A union job was a test away.
OK, Iloved my father, but didn't want his life, including the drinking, and Archie Bunker mentality.
I ended up working my junior year in highschool, and going to College of Marin. Back then it was called the Little Berkeley. It was a great school for many years. Those two years were with the cost!
I wanted to become a doctor for all the wrong reasons, basically I could memorize visually all those biology charts, and organic reactions. It just came so easily.
I needed a 4 year degree though. I went to --- ------- state for the four year degree, and graduated. I also had a very aggressive girlfriend at the time, and I knew she would hold it against me if I graduate from that joke of a school.
I had a panic attack while I was in graduate school on december 24th. It was probally comming. I was very much a hypochondriac at the time, and actually believed I has a brain tumor. I tried to go back to school every month for a year, but was just a neurotic mess. I went from being the most capable guy in the room to not being able to drive a car.
I was so neurotic. My girlfriend was a saint though. I had this period where I though sex was making my head pain worse.
We are talking about a 19 blond virgin until we became a couple. She believed my bullshit until she saw a Woody Allen movie with her mother. I belive it was Hanna, and her Sisters.
All I knew was I had this head pain, and life seemed to short. I had a hard time repressing my sexual desires. I had a hard time sitting in school--delaying gratification. I did know that the worst day in college was better than any job blue collar jobs though.
(Why did I write this? I use HN as. journal. My respect for this site has dwindled. When Dang isn't hellbanning me for no reason------I will write. It's not for anyone's benefit except myself, unless it comes to real subjects. I am not going to edit this, especially after a bottle of wine.)
Too many times, the class was supposed to line up at one side of the gym, then one at a time each kids was supposed to do the thing.
I was never good at waiting in a "useless" line.
I have a more mature perspective now, you never know what struggles people have and the courses are really tough for a lot of people.
>
> I have a more mature perspective now, you never know what struggles people have and the courses are really tough for a lot of people.
Well done on you! I occasionally run into past versions of you. They usually dismiss my BSc degree and my MSc degree with an attitude that is similar to:
"I learned real lessons at the school of hard knocks. While you were partying with your college friends and memorising useless theory, I was making ends meet and learning practical programming by actually doing it."
Truth is, I've never been to f/time university. After school I (very briefly) apprenticed as a auto mechanic for a short while, then left to work in a factory assembly-line (12-hour shifts, all night-shift, 7 days a week) for a little more peanuts than an apprenticeship paid.
I used almost all of my meagre income to pay for part-time university (work at night, study+sleep during the day). Halfway through my second-year courses I finally caught a break and got a job as a computer-lab assistant at a nearby university.
It wasn't actual programming work (show new students how to log in, refill printers with paper, help students who destroyed or lost their access cards, etc), but it left me a lot of free time to waste on usenet, which is where I found my first actual programming job.
I don't narrate my origin story to those past versions of yourself, though. There's no point. Their self-identity includes their own bootstrapping story about how degrees are pointless.
However, I do completely agree there are other great programs out there. UIUC and UTexas have great online MSCS degrees that are comparable and well worth considering.
The coursework in OMSCS can vary. There are ways to get through the program taking fairly "easy" courses, and then there are much harder ones. I have certainly had a class or two that was less useful to take than others, but that was mostly so I could graduate in 10 semesters as opposed to taking longer.
Education is an interactive experience, the value of a good professor and classmates is helping to identify issues when you are stuck and nudge you just enough for you to connect the dots. As an undergraduate I liked attending study groups and office hours. I learned as much from my peers, both me helping them and them helping me, as I did from my professors. This spirit was difficult to replicate this in OMSCS.
The problem is that high tough, in-person communication is very effective but this doesn't "scale" well. I can only imagine how different the world would be if Plato was teaching online. :)
Also I was in the 2nd cohort, there were certainly a lot of "bugs" and issues they were working through. Additionally, the culture of video conference and zoom was also less a few years ago, which could make an impact.
But there is still little to no interaction with the professors.
I actually found the interactive discussion to be more frequent then my (in-person) undergrad
Slack and other conferencing tools are game changers. People getting used to these tools is one silver lining to the pandemic.
How much time do you invest every day on this degree?
I dropped out of my MS degree 20 years ago and would love to do that again, but I feel like I would only have (at most) 1 hour a day to work on it due to all my work and family commitments. I could squeeze an additional hour at work to learn but that's about it.
Most people either do a couple hours an evening plus a chunk on weekends, or else dedicate most of their weekend. I tend to fall in the first camp, and then also add more weekend for harder classes.
The time commitment varies from 5-30 hours per week depending on the class. Most classes fall pretty close to 10 per week in my experience, but I also have been mostly picking medium difficulty classes on purpose.
Source: graduated from omscs last year, with two small kids and a full time job. Took 3.5y though, and I had to skip a couple of hard but interesting classes because they would have been too difficult given the situation (covid lockdowns with kids at home and increased job-related workload)
The old adage "where there is a will there is a way" is definitely true.
I completed this at one class per semester, except for the semester I attempted two classes to speed up the process. While I was able to complete both classes, I felt the learning suffered as I was just rushing to complete assignments in time instead of really learning the material.
If you can accept those sacrifices, and your partner is willing to support you (mine wasn't, it turned out), then it's worth considering. I would also strongly suggest treating it like a marathon, not a sprint, and do the 2 courses a year option.
I did it mostly because it is easier to get visa
Generally i do not consider higher edu institutions as a place where i can do/learn advanced stuff related to computers
But you could say the same about high school. It doesn’t follow that reading writing and algebra are bad vocational investments, in fact the opposite because they support many possible vocations.
As a CS professor, my goal is to push students to rewire their brains in ways that are better and more effective. It is also for them to gain foundational understanding that will allow them to learn advanced vocational skills. I teach theory so software engineering may be different. But of course I don’t want to teach something that will mainly be useful only for the next 2-3 years of their career, I want to have an impact on the next 20-30.
In the future I'm hoping to run their tests inside Valgrind to check for memory leaks, enable the various sanitizers, etc., and I also need to figure out a way to check for conformance with my code formatting standard, but for right now it gives students perfect knowledge of what their grade will be when they submit.
That said, I spent a lot of time building extracurricular websites in my spare time, which was also a huge leg up in my future career.
It's still not what's most important at uni, imho. You have 5 years at uni, many decades at your job.
The power of a university education is on several levels. One is the community of peers that you are part of. Second is the ability to learn what you need to and to learn what the limits of your knowledge are. Tech changes quickly and no university syllabus will ever be fully up to date. Besides, that is never their goal. Third is to to provide a safe environment to try and fail. Finally, a university degree is a basic filter. Completing a degree shows an ability to learn new things, be tested and demonstrate a decent understanding. You may fault the methods or the result but the process is effective and no, bootcamps or community work are not a substitute.
A university is not all about research but research is an extremely important skill to have. it manifests in ways that you may not imagine at the time.
What???
It is incredibly stressful environment cuz if you fail then youll have to spend more time (so less vacations), maybe even pay some $$ or at worst have to take course again
There is time to learn that your habits need to be changed in order to be successful in the long-term.
There is more support for investigating things that you think are worth investigating, and there can be positive reinforcement that some people find meaningful (e.g., repsect for presenting at student conferences) instead of "just" a paycheck.
University certainly provides a way to establish your credibility and build a portfolio that can be broader than $dayjob gives you the opportunity to do.
At the university there is very easy access to people with a very broad skill set. No need to beg HR to allow you to hire an ML specialist... just go down the hall and talk to Professor X.
It's not perfect, and not everyone makes as much of it as they could, but there are big plusses to the university system.
In my world there was always lack of time cuz other courses wanted their important^tm projects too, so instead of deep dive into fancy topics that was just pass and move on and learn it yourself
In hs i liked math, in higher edu institution i had no time for it
But ive been working full time almost whole time
I carefully evaluted every programme. I spent probably a year or two looking at different programmes, not through analysis paralysis, but because I was looking for programmes that had things that interested me and weren't either simply a shake-down for money or getting lost in proofs and theorems. I do better with project based work that builds up, rather than term papers and exams, which I am terrible at. I know myself, so I picked appopriately. My wife is the complete opposite.
My focus for picking a programme is always "show, don't tell" when it comes to "how do I demonstrate my knowledge in this subject." It is what works for me. Pick what works for you.
There are a lot of good master's programmes out there, but you really have to do your due diligence and not just rely on some website that screams "100 best online master's degrees!1!"
Look in to Open Classrooms, Thomas Edison, Structuralia, and several others just as a starting point. Don't get lost in the rigid academic weeds of North America. There's an entire world out there to explore.
It comes down to "are you looking for academic rigor that leads to more academic qualifcations later" or "are you looking for an education that will serve your professional needs." Most programmes are the former, and few are the latter.
I spent about $10,000 on my most expensive one. I spent 3K on my cheapest one. There was no difference between them in terms of what I got out of them.
OpenClassrooms also does EQF 6 (Bachelor's) and EQF 7 (Master's) diplomas, some of them are backed by Universities in France (Ecole de Management, Central Supelec, and others), and again, may be a degree or may be a diploma - but if it is a degree or diploma backed by a University with a charter, then it is a real qualification. Same goes for Structuralia. You need to check the programme. At the end of the day, if you have an EQF 6 (Baccalaureate) or EQF 7 (Master's), and it is underwritten by a university with a charter, you're golden. It only matters if you need the thesis for going on to a terminal degree (and even then) or want to be in academia. I don't. I'll take the five master's and be happy with it. If you do any of those courses, you're still going to get put through the wringer, it ain't a walk in the park. I'd say the AI one, though easy because I knew the subject pretty well, was the hardest because I really stretched far beyond what I had grown comfortable with.
I have studied at nine different European universities/colleges at various levels, one in Asia, and four in the US. I feel I got value out of all the education (except that one MBA with a concentration in HR that I absolutely hated), but I prefer the European system most.
Footnote: When my wife mentions my qualifications in conversation, I correct her and say "two master's degrees, three master's diplomas." I still have that conscious bias myself. But every time I check the paperwork and every university I've shown them too for my next educational step, the university has come back with "that'll do nicely. These look good." It's that impostor syndrome kicking in again.
And one more thing...
Bachelor's and Master's with and without thesis doesn't delineate degree or diploma. It comes down to how the certificate is awarded, not what you had to do to get it. My Master's degree in Computer Science was a "without thesis option." Education above grade-school gets muddy, which is why you need to do research on the subject.
You can earn an entire Bachelor's by testing out and/or doing prior learning assessements (I've thought about that if I could find a subject that interested me). PLA acceptance is far more prevalent in Europe than in the US, though Thomas Edison in US does let you test out or do PLA for the entire degree if you are capable of it. I'll point out that doing PLA is not a walk in the park. I've done a couple of prior learning assessments where I felt I knew the class subject really well - it was probably more work to do the assessment (lots of writing and gathering of research notes and documentation) than it would have been to just finish up the prescribed projects.
I am no expert in any of this, and I strongly recommend you do your own research.
Only when you get smacked by an exam or an assignment you realize that you've been fooling yourself.
The ML and Systems track is pretty hardcore. If you survive those you have pretty good foundational knowledge.
Fair warning though - the program is not a joke. Georgia Tech always leans towards application. You will spend a lot of time writing code and tinkering, but it's worth it if you want to learn a ton.
Have your classes been built around group projects, or can you go it alone if you're doing it for kicks? I'd love to spend some time studying CS and end up with something to show for it, but I'm not interested in group-work-as-workplace-prep.
The key is to pick classes that are hard, and form groups early. People that reach out earlier usually have their shit together.
I mean, if you want to do it just for fun, knock yourself out. Just don't delude yourself about what you hope to accomplish.
I can see a lot of people wanting the job of "play with cool toys and don't be responsible for products", especially as a mid-career transition.
SeparateLy… I find the same line gets used both for and against this degree: “the Masters is the new Bachelors”
FOR: to say the Masters is now the expected baseline — you need it like you used to need a Bachelors.
AGAINST: to say it’s already been watered down and/or too much quality variance - that it gives about as much signal as a Bachelors.
That only applies for very early career. You get to skip one to two of the most junior levels in the company. By mid career, there is no skipping anymore
There is a bit of a leap from more likely on a low probability event to probably will pay off their debt. Getting promoted at a Faang company takes a good bit of work. Becoming a director is not something most people can expect. I’ve never heard of a promo committee at a Faang Corp take a degree into consideration and I’ve known plenty of director level people at faang companies who had no grad degree. If the person learns some useful new skills that could open doors. The credential might also help open some doors that an obscure college would not. But saying they are likely to make it back in 6 months is a bit of a stretch.
If you don't think a lot of places aren't gatekeeping on degrees, you haven't looked at many.
To be more clear for folks that think it is an absolute filter, of course it is not. I remember about half of hiring pipelines ending on that point last time I was looking. Was 100% of the top companies with too many applicants however.
Company A: Lead (yes, but undergrad in EE), Manager (no), GM (no), VP (?), SVP (no), President (yes), CEO (dropped out of MBA)
Company B: Lead (no), Manager (yes), Director (?), Senior Director (no), Senior Director (yes), VP (no), SVP (MBA), CEO (yes)
The plural of anecdote is not data, and I'm not a director, but a cursory comparison of senior leadership and LinkedIn doesn't leave me convinced that a graduate degree matters. Most of the folks who have one got one back in the 90s where things were very different.
https://www.aplu.org/our-work/5-archived-projects/college-co...
This one says 400k+ higher lifetime earnings with "advanced degree". The source from georgetown says they are less likely to loser their job in a recession as well.
No, the real world is not black and white and this is not a guarrantee. But, get an advanced degree looks like an easy power-up at the prices OP was willing to pay.
I can only think of aerospace, and other engineering-not-software companies that do this. They also don’t pay sw developers much to begin with, so this looks like a very false economy.
Not true at all. Most FAANG companies factor in your MS in you compensation (especially if you are moving from one to another)
Also, after a career in startups, I am looking for work in bigger, more conservative organisations. So those are the places I want to work.
OTOH, UT Arlington is no Stanford, so maybe it's worth it if you can get into one of the top 10 or so programs in the country.
So at the end of the day, whether a master's degree is worth it, I think depends on whether you've got a good enough foundation already - if you have that (from a bachelor's degree or otherwise) then a master's degree only really makes sense if you want your career to target the kinds of companies (typically enterprise) where where there's a bias toward promoting mostly people with master's degrees.
In OPs example, they have a bachelor's degree in physics, so I think it makes sense to add a couple years of computer science on top.
Masters is just for degree
In many places it's a matter of employability. Without a degree, you are simply not employable (or promotable) unless you have the right network.
Discussing the relevance of the contents of a master's is like discussing the relevance of leetcode. At no point does anyone, either from the university or industry, claim the contents are relevant to the job. It's also why many degrees are effectively interchangeable as long as they are similarly prestigious.
I think mid-career degrees can have the effect short-circuiting the elitism.
I've been in the industry for many, many years now, and the work I do has a small amount of overlap with the contents of a CS degree. But a lot of software engineering in the real world exists "above" the layer that CS teaches, not unlike CS living above electrical engineering. On top of that, in general, soft skills seem to matter so much more than having 10-30% more hard skills. This varies with your position of course, but I'd wager that the majority of software engineering positions are like this.
Having a credential from a decent school might help get more compensation, but without being able to truly quantify the amount of "help" it gives, it becomes impossible to adequately quantify the ROI.
And sure, you'll see some job postings from time to time that say they "prefer a masters degree," but what about once you're in the door? I don't think I've ever heard a software engineer explicitly put much weight onto the educational credentials of a peer. There may be unconscious biases, but for most engineers, it's about producing results.
I don’t have a masters. I’ll likely never need one. If there was a specific field that required intense learning and was cutting edge (computer vision or something) then I’d understand a masters (or especially a PhD) but for general work…? Don’t get it.
You’d certainly be better off financially from doing more leetcode or system design prep instead.
Marrying software development knowledge with another non-tech domain expertise would make one valuable.
To who though? A lot of people won’t care if your masters is in education if you’re gonna work in most fintech.
I think it pigeonholes you further if you go down this route. If you want that then go for it but I’ve seen many wasted masters and PhDs in Silicon Valley. Many.
"To whom..."
Sorry. Couldn't resist...
Muphry's Law strikes again: compl-e-mentary vs. compl-i-mentary.
The masters students at my current uni (ETH Zürich) are easily the best software engineers I have ever worked with.
If you don't have a goal like that, and you already have that level of experience, I agree it's dubious what a masters would do for you.
This would only make sense if you're looking to retire-in-place though.
I've done almost all of my learning either on the job, or through extracurricular reading (e.g., lots of conference proceedings, and textbooks on interesting subjects). If you read just one paper a week, that's 500 papers in a decade, and you'll have a good grounding in large swathes of CS (or at least you'll know where to go for ideas on hard problems). Two papers a week and you will be wise beyond human measure.
I dropped out of college 40 years ago and the lack of a degree has never been a problem. It's getting that first job that's probably the biggest hurdle; after 4-5 years, few companies will care. And if they care, you may not want to work there.
I've seen requirements for a master's in CS or related disciplines on job descriptions for technical leadership roles, mostly at larger, older, more conservative organisations.
My work experience gets me roles as director of eng, vp eng or CTO at most startups, but for the national rail company here in Canada, I have a hard time getting considered for any jobs.
For two cents, I’ve been in leadership roles (VP, or VP-equiv) at three big/huge companies now (Oracle, Amazon, Microsoft) and I don’t even have a Bachelors degree. This is somewhat unusual but not a one-off by far.
Job descriptions often say “or equivalent experience” (meaning professional experience) and in probably 1-2 cases during my career, a requirement to have a degree with no “or” clause included still hasn’t been a huge issue.
Impressed with all the research you did here and and your being intentional about objectives and how to get there, good luck!
I’ve hired folks at all levels. Degrees are one data point among many when I decide to interview a candidate, rarely discussed when making a hiring decision, and _never_ a factor when making a promotion decision.
If considering formal education to unlock promotion, I would strongly suggest speaking with one or two people with recent, direct experience in the promotion process at your institution. Get familiar with the functional reality of the process (it’s usually more subjective than advertised), ask how credentials are used (they usually aren’t), then go from there.
I will say that I have required education as evidence of commitment when considering applicants who were looking to change careers (that is, I see you have had success in your current role, but go get a cert to demonstrate basic aptitude and show me you are serious about doing this). This was only relevant for the most junior, entry level positions.
As an aside, I’m currently pursuing a teaching credential through WGU, but I have zero expectation that it will provide additional income (quite the reverse!) Rather, I’m pursuing the credential as a forcing function to engage with topics outside of my professional experience, and to provide access to opportunities that are actually formally, legally gated by credentials (teaching in public school in California).
If so then I understand - if you've seen other people getting promoted and wondered about whether it was the extra qualifications. I am not really sure at the moment. I used to think that way and then I got a promotion which was nothing special to anyone except me - to me it was a huge breakthrough. It was not just because of my personal performance but about how extremely useful I was to my boss at the time and because he had another problem to solve which he thought he could put me onto. In other words it was all circumstances that mattered. To be fair, it is also because he is probably the best manager I've had so far (he thinks that hiring people who are nice is more important than anything).
Before my mini step-up I had been in companies where either the source of all solutions was always expected to come from outside i.e. we were to be soldiers without initiative or situations where other, better connected people were in the view of the management. There wasn't any chance for the people 2-3 levels above to form any opinions of you and the person directly above you was basically a competitor. I'm not joking that "the smokers," who had to leave the building to get their fix, was the most important clique to belong to at one place because people of all levels mixed there and got to know each other.
To get back to your Masters, I think it's a heavy investment for someone whose top concern is advancement/prestige. It feels a bit not right that the things that you want to learn about aren't at the top.
It might be that you will do all this work on learning things you're not really that interested in when it could turn out that the real problem is that you're just not working at a place where the conditions are right to show your usefulness and/or you don't have good management above you.
It's 2 years out of date, but might still be useful:
https://docs.google.com/spreadsheets/d/1mCETBcQJ5V1UA2maojCa...
In the end I didn't do any of them because they're less productive than self-teaching (basically YouTube + Kaggle + real personal projects), and I don't value the CV cred highly.
This is the most important line in the article. Perhaps that’s part of being “mid career” but that’s an excellent level of self awareness and shows that there are multiple upside/benefits.
I wish him the best of luck!
Was it hard? extremely. a lot of concepts were very old in my memory, but it was also a lot of fun. I enjoyed going to class every day.
If that's local, affordable, and easy, I wouldn't worry too much about looking for something "fancier" right off the bat, I think there's a lot to sink your teeth into there. There's a lot of stuff online too on all these topics of course, but if you're like me, that's not the same as taking a live course.
My recommendation is don't waste your time and money unless you have a specific goal in mind or want to go on to a master's program - which I also generally don't recommend unless, again, you have a very specific goal and no other way of accomplishing it.
Data Structures and Introduction to Algorithms: everyone needs this and it is maybe 50% of leet code problems
Assembly Language Programming and Machine Organization, Introduction to the Theory of Computation: this is the low level of how computers really work
Operating System Fundamentals: concurrency programming plus a lot more useful stuff
I think it is definitely worth it. Good luck!
What’s the other 50%? I’d have thought DS&A would cover everything.
I am not sure if there are any problems on leetcode that require any of these, but based on a sample of a few of their contests, if they exist they are a very small fraction (<<50%) of all problems.
The vast majority require:
- no DS&A knowledge whatsoever
- basic techniques (recursion, simple graph/tree algos, binary search, simple dynamic programming)
I completed all the "fun" courses, the ones that really taught computer science, but the focus and attention I needed to complete courses like advanced logic and discrete mathematics III I just couldn't combine with a fulltime job. I'm no math wiz, so those courses were really tough for me.
I now realize there's other universities that have less maths and more of the fun stuff, but even still say I wanted to jump to another specialization if I had 2 full time years to do it I could spend my time so much more efficiently than at a university.
The "not fun" ones on the other hand are the ones I remember the most from because those memories aren't competing with lots of other exposure to those concepts and thus they left a lasting impression that I might not have gotten any other way.
Going down the path of "fun" might be easiest and overall more enjoyable, but it also ultimately might be less fulfilling in the end. Your mileage may vary, of course.
OMSCS has a lot of flexibility in being able to attenuate your course load, gives 6 years to complete the degree, and your peers are mostly working professionals. That's a big advantage of online-first programs.
I got some mentorship that triggered growth on an intellectual, professional, and even personal level. At the end my advisor really worked hard to put me in a great job that was exactly right for me, it didn't change my earnings but the work is fascinating and the environment is excellent.
All this to say that like anything else it is what you make it and it's all about specific relationships. Look for specific faculty members you might connect with, even just taking their classes.
Also if you do enroll, make sure to go to campus at least once a semester: shake some hands, sit in some offices, make sure people can put a face to an email signature.
I've also seen people mentioning if they were going to get a masters they would do so in a different field. I just finished one in policy. I learned much more in this masters than I expected. I realized there are many games to be played, this article was posted to the top page this week and resonated because of it https://news.ycombinator.com/item?id=31267685
Link: https://statmodeling.stat.columbia.edu/2022/04/24/more-news-...
#1. The fundamentals are the same, but a lot of things have changed in the past 30 years. The coursework, somewhat, brings you closer to up-to-date.
#2. Non-FAANG. Looking to be a director or a tech fellow? No formal education during your career is a bad signal. Of course there are exceptions, but it is expected that you have invested in yourself (MBA also meets this requirement).
My personal achievements:
#1. Greatly improved my writing skills.
#2. Understand deeply and can communicate technical concepts that previously I dealt with on the "gut feel" level.
#3. Able to identify ignorance and stupidity in my leadership directly, where previously I lacked enough context to be sure of them being wrong.
Is it a cure for imposter syndrome for those who think it's due to not having a CS degree? I don't know. It's a very expensive cure (in time and money) if so.
Basically anywhere where you have to compete with a lot of people, it is the first filter to cut the stack from 100 to 50.
I’ve seen data scientists from political science backgrounds, UI engineers who studied literature and philosophy, and I have a CS degree myself, but I’m originally from the bioinformatics space.
Now, networking is a thing of course. Many people get in that way, but for J. Random Applicant reading this, not an option.
Interesting list of examples, I wouldn’t necessarily have grouped these together?
Have you applied for these at some point? I haven’t, but the emphasis seems quite different at each company
This is the ONLY program I'd like to get into because so far it is the only program that I know deals a lot with hardcore malware analysis. There are probably more programs available but I didn't spend much time researching options.
Unfortunately this program is NOT online, and I'm not in Europe. ALAS! If anyone knows a similar online program in NA please let me know!
Me sending application of withdrawing a few grand to $wifie
But I "get" it. I've thought of doing this too.
I would add one criteria: I get to do the whole master's degree using Rust.
I'm sure this would narrow the field of schools.
I want a master's from a school that recognizes that it's effectively over for C++ and Java. At least it's completely over in my own mind. There's no way in hell I'm doing a master's degree where the school's obsolete syllabus dictates those legacy languages.
This stipulation might garner considerable traction. This quirky old guy, with more industry experience than anyone on faculty, will run advance reconnaissance for us, which is something he's been doing much longer than our typical master's students have been alive.
Because let's be honest guys: it's all about Rust going forward. If you feel otherwise I respect that view, but you haven't seen the shit I've seen.
I'd do a Rust-centric master's degree in a heartbeat, and I'm sure I'd love every minute of it.
This applies not only to languages, but techniques, systems knowledge, documentation - everything about it really.
Part of the problem with selecting a degree on the basis of modernity is each course in the degree can be vastly different - some are up to date, others are 10 years out of date.
There's also a slight risk in doing a course in its first year, because the bugs haven't been ironed out.
One of the most famous introductory CS courses in the world, MIT's "Structure and Interpretation of Computer Programs", was for decades taught using a programming language called Scheme, a dialect of Lisp that's wildly different from any mainstream language and is used for basically nothing other than this course. This was partly intentional, since it does a great job of shocking teenage hacker wannabes out of their "I already know language X so I'm hot shit" complacency.
At the beginning, you're a 20-something person. You mostly don't get to pick your steps, especially if you're in the employed-by-others track.
Standing in the river on my rock, at this advanced career stage, I'm certainly not looking to "learn programming languages". That was long ago.
Right now it's all about picking the next leap based on the conviction that it's going to be a great ride in the future, which is much closer to sunset for me than most in this forum.
I'd do a master's degree because I'm curious as hell, and having tons of fun learning surrounded by smart people is all I want to do now.
Ah, you fell for the trap. Notice how the course description you link is hosted on purdueglobal.edu? Purdue Global (formerly Kaplan) is totally unrelated to the main campus. Basically they took a garbage online college and slapped the Purdue name on it. The “real” Purdue of course has an MS in CS program, though AFAIK it’s only full-time and in-person.
Oouf. Felt that one (and I don't even have a MS degree).
I'll share a comment that I made when I read that article: please note _technical interview performance_. Not job performance. Because a MS may not teach you leetcode-style algorithms as a BS. Spoiler: OMSCS does, with graduate algorithms, which is in fact quite pointless if you're a mid career software engineer, but useful if you're a career switcher.
The SE program at Oxford doesn't seem to be available for 100% remote study, though. Am I missing something? I love this informal, friendly Web site, and I might be looking in the wrong places.
What was your background when you applied and how tough is entry? (I see that there is a 99 application average over the past three years with 70 acceptances?)
And how much is the cost?
You get a degree from Oxford at the end, which definitely opens doors. Of the ten modules I did, 2-3 were exceptional and stay with me today. 2-3 were trash. There rest were OK. It's an excellent excuse to get familiar with the subject, and the one-week onsite format followed by the six weeks of coursework is excellent.
> would you recommend it for an American? Is it even possible for an American to fly in for the week-long intensives like the other comment on here says?
Yes, we had quite a few Americans in my cohort doing this
> What was your background when you applied and how tough is entry?
I essentially flunked out of high-school, but I had ten years of solid commercial experience under my belt, and some interesting open-source contributions and tech talks at conferences, and they let me in. I suspect I had the potential to interview well as I also was awarded an academic prize at the end for highest overall marks...
However they also had this really interesting system when I was there where you could essentially earn your way in: they'd let you come and do the SEM module (challenging, but not nuts), and if you passed, they'd let you in as a real student.
> And how much is the cost?
£41,080 over four years for a non-UK resident
I don't know how tough is the entry because I have not seen the other applications. What I did appreciate in the application process is that since the target audience is working professionals, you can get 3 letters of recommendations from people that you worked with, not necessarily academic recommendations (for someone who has been out of uni for a while, makes the process much less painful).
Some other selling points of the programme for me:
- 1-week modules, instead of one evening per week - much easier to organise if you don't live in Oxford
- No exams - for each module that you attend, you have one week of lectures, followed by an assignment that you have 6 weeks to submit the answers. Assignments demand a lot of time to complete, but I find it much more manageable when you are working than cramming for one big exam
- Flexible curriculum - to complete the course, you need to complete 10 modules plus a project (like a Master Thesis) in 2-4 years. But you have complete freedom to pick and choose any course from those that are offered (http://www.cs.ox.ac.uk/softeng/courses/subjects.html). So you can customise the programme according to your interests.
- Time flexibility - you can complete in 2-4 years. There are 3 terms per year, so you can attend one module per term and finish in 4 years, or if you have more time available, attend more modules in a term and finish in less time.
The downside of the flexible nature of the programme is that you have less of a feeling of "class of '19", or whatever the year of your intake. You will bump into people that are taking their first or last modules, and some of them you will only see once. You will still have the chance of going out for a beer with them, so I don't mean that you won't have a chance of networking.
- It is in Oxford, which is a pretty cool town to visit, and as a student you have free access to places like this: https://en.wikipedia.org/wiki/Duke_Humfrey%27s_Library#/medi...
I’m curious to hear about the SEM (I assume Software Engineering Mathematics) method of qualifying perhaps as a post-graduate (non-matriculated) learner — this taken from the other commenter. This seems like an ideal way to test out the structure and style of a classroom and qualify by achieving high marks before committing and fully applying to the program. Is this a thing that you’ve heard about? There isn’t anything on the website about qualifying for the program through this.
Also, what was the application interview like?
> an ideal way to test out the structure and style of a classroom
My understanding is that all the courses are explicitly available to anyone who wants to take them and can be booked by members of the public, although expect there to be priority to matriculated students for very popular courses (of which SEM is _not_ one). They were about £1,400 each last time I checked.
I don't think I'm giving away anything I shouldn't by saying that SEM is the first 10 chapters of http://www.usingz.com/ taught in classroom form (see the course desc at: http://www.cs.ox.ac.uk/pro/subjects/SEM.html and compare to the TOC of Using Z), and certainly I benefited by having read and understood it first (which was very challenging, but also SEM has kept coming in useful at the weirdest times in my professional career).
The other take-away is that if you don't have the math pre-reqs for a course, and you figure that's okay because you've always been pretty good at math & you'll just power through: don't. Take the pre-reqs first, unless you want to needlessly suffer.
Honestly I would love to get another MS in data science and/or statistics. I could go the self-learn way but I need the structure and accountability that is required to obtain a degree. Otherwise I get side tracked in my self-learning path ... SQUIRREL!
I would like to know How are a program like this valued by industry / society? Whether it's valued or just skipped by many since it's "not real/online" master degree.
Full online courses like this interesting for me who's based on the other part of the world.
For me it's a way of focusing on some topic related to my work: I'm a machine learning engineer and I'm doing the PhD in AI/optimization. I have spent just a year on it so success is still far away.
I know people however that went back to academia but they had some previous experience for example as teaching assistants.
Disclaimer: never been in industry, but knows several people (PhD students, professors) who were in industry once
To the commenter who said that a masters won’t help you and is only for entry level positions… that’s simply not true. Many positions have a masters degree as a prerequisite. Having an education is never not a good idea.
At least in the US this is not true. Many people on the leadership at Amazon and Google (at least in my tenure there) didn't have masters or PhDs, what matters most are your professional achievements, actually smaller/medium corporations are the ones that will be impressed with your cert/degree. I have seen managers with no degrees( that started as a self-taught developers and later became managers) managing engineers graduated from MIT/Hardvard/etc. Maybe in Canada things are a bit different in that regard. Not discouraging OP or anything , everyone can spend their money however they see fit. Just pointing out of the myths I see often in legacy media.
Good luck with compilers...
It looks like the GRE is now available online, 24/7/365, so I could probably study and take it in the next month or two. There were a cluster of Fall 2022 application deadlines at the beginning of May, but if I decide to apply for Spring 2022, a GRE might be worthwhile to give my application package a boost.
Thanks for the note!
Its 2022. Been 10 years in the industry. I cannot agree to this.
I clearly see Big companies (recruiters, MANGA) selecting experienced resumes based on years of past experience, resume keyword matching, leetcode expertise and salary expectations.
1 year of leetcode + 1 year of AWS positions me better for an offer from Amazon than a Masters degree.
Different if you want to upskill as a programmer
Unfortunately no, not really.
If you've got some data to the contrary, please feel free to share it.
But it’s also hard in the UK. Maybe there are more universities that allow it but don’t advertise it anywhere but the number who will state clearly they will take work experience in lieu of a Bachelor’s is small.
I’m in a similar position (incomplete BSc, but 20+ years commercial experience) and was looking at the OU’s masters but wasn’t that impressed with the syllabus.
Examining your motivations is not easy and should be done with someone that knows you well, isn't family and can offer objective/'cold' advice. If they fail to dissuade you w/some real life experiences and you are still fired up, then you should probably go for. it.
TLDR do not embark on a major undertaking without due process (or self process, as it may be).
Trust me, I've 20 years of software engineering exp and an undergrad in CS.
It doesn't meet my requirement of having a real master's degree from an accredited university without the "extension" or "electronic" qualifier, though.
I'm doing the Algorithms specialisation from Stanford on Coursera right now, just to make sure I'm ready to get started on a master's in the fall.
Well...it wasn't as educational or as fun as real world school. Those who claim Discord is a substitute (or even better?) for in person collaboration in an academic environment with like minded peers, uh ok...I am in those Discord channels and...100% disagree.
One thing I found out about Georgia Tech's graduate program is that it's more of a measure of grit, not a measurement of your knowledge. Their harder courses aren't necessarily hard material that one can't grasp. It's that the 40 hour projects...it's 10 hours of learning/implementation and 30 hours of a few little nuance issues that add no value to your knowledge base.
If you want to see how difficult the material are - many of the courses are on Udacity. Projects won't be listed, but maybe you can find similar ones online.
Since my employer pays for education (not just graduate, but if you don't have a BS they'll pay for it too!), I have a lot of colleagues as well as peers in industry who went the Georgia Tech route. It makes sense. Pure online, inexpensive, renown.
Everyone I know - and I mean everyone...when they have only a few classes left take the easiest ones to get it over with , i.e. Digital Marketing and whatever is ranked easy on OMSCS Central.
If someone is young in their career and wants to upgrade their piece of paper this would be an OK route. If you're older - you likely have an idea of what you want to learn. And if not, well, just watch all the Tech Udacity classes! There are plenty of online resources and communities of top professionals to surround yourself in where your time is spent more effectively.
As for me - I found myself taking Digital Marketing for my 5th class...I dropped my 6th class after getting tired of a particular bulls--- project. I will wait and see if I decide if I reenroll in the Fall. I am leaning towards a no.
IMHO if you are or want to be a software engineer, plenty of more relevant resources out there to grow your knowledge without wasting your time. If you want to be a hard core computer scientist, do an in person program so you can focus and be surrounded by peers and professors.
Tech's OMSCS is ground breaking in that it provides a quality education to everyone. But it is not a quality education for everyone and I wanted to have at least one post here with a contrarian point of view.
The challenges I've seen with masters programs fall into a few broad categories:
First, few admissions processes seem to really be designed to accommodate returning students, especially those who have been out of academia for a long time. Most schools want recommendation letters, and expect at least one or two of those to come from academia. You can get professional recommendations, and it seems like having one or two is a benefit, but a lack of any academic recommendations seems to put you in a bad position for admissions.
It's also quite hard to make up for other admissions blockers if you've been out of school for a while. If you have missing prerequisites or a low GPA your out of luck. Finding out you need more research experience seems to be another significant blocker, although I've heard of people in industry successfully volunteering time working with researchers to get experience ahead of Masters applications, so that could be possible in some circumstances.
Overall, it seems that unless you finished up your undergrad program already planning for a masters degree, you're going to be stuck with some semesters of taking undergrad courses as a non-degree seeking student to fill in missing prerequisites, get the relationships with instructors that you'll need for referrals, and possibly get documented evidence to counter questions about a low GPA.
That ties into the second problem: Most programs lack the flexibility they'd need to provide real value to experienced developers. Masters programs could have a lot to offer experienced developers without a formal CS background, but they are largely designed for people with little work experience, and the requirements involve a lot of courses that are going to be a waste of time for someone with substantial industry experience. The vast majority of masters programs have some required courses designed to prepare people for industry that are going to be superfluous to anyone with substantial experience. If you have a lot of missing prerequisites on your transcript that can be even worse as you're also required to take unnecessary courses outside of a degree program to even be considered.
The lack of flexibility also manifests in another problem: The abysmal state of online programs. There are a lot of bad masters programs out there, but even the good ones aren't going to be good for everyone. Every school has a machine learning and cyber security masters track, and if you have interests outside of those areas then you're going to be substantially limiting the already narrow choices of schools that have a decent curriculum that you can get admitted to. At least for my areas of interest, few of those had quality online programs. That said, suppose you do find an online program that aligns with your interest, the online education system these days requires complete submission to a ludicrous panopticon of the shittiest and most invasive software imaginable. If you like in-person learning or you don't want your life for the next several years to be massively and shamelessly exploited by proctoring spyware, then you better be prepared to relocate, and hope that your manager is okay with you being gone in the middle of the day a few days a week, because few of the best programs offer evening or weekend course options.
After looking at the state of masters programs, I personally decided the easier thing to do was to simply continue self learning, making contacts with researchers who are working in areas that interest me, and work to establish myself enough to have a chance to collaborate independently on research and find ways to get into the kind of work I'm interested in without formal credentials.