9 ms·
I was having another one of those moments yesterday where I was questioning the value of higher education while reading a HN thread about teaching yourself CS.
by dougmwne 6y ago
I was having another one of those moments yesterday where I was questioning the value of higher education while reading a HN thread about teaching yourself CS. There was a healthy debate about whether the typical CS curriculum helps a modern software engineer at all. Some thought the degree was stuck back in the 70's. Others thought there was value, but had trouble articulating it. It seems that when the information received was divorced from the credential, the value dropped precipitously. And this is a hard sciences degree with many "universal truths" to teach. Things get even more squishy when you get into the humanities which are much more idea, opinion and discussion based.
As the linked article points out, employers often look for the degree, but do not believe it is much of an indicator of quality and don't expect new grads to walk in the door with the needed skills. As a former recruiter I think that sentiment is pretty spot on. I don't think I ever asked a candidate about the details of their education. It was just never relevant in comparison to their experience.
So that leaves the on campus experience, which I personally think is amazing. It's like living at a country club for 4 years, crossed with a huge expansion in your world from living at home and going to high school. The new ideas from the classes are just a small part of the new ideas you get exposed to from other students and campus events. Plus with the increasing independence, it's a great gentle introduction to adulthood. Unfortunately all that is not available during the pandemic.
So the content of the classes is in question, the value of the credential is in question, the cost is ballooning and the on campus experience is on an uncertain hiatus. If ever there was a time to disrupt higher ed, that time is now.
- beojan 6y ago> whether the typical CS curriculum helps a modern software engineer at all The purpose of a computer science curriculum isn't to train a software engineer, it's to train a computer scientist.
- dougmwne 6y agoYes, I absolutely agree, which is why the popularity of the degree is so utterly curious. Companies do not need many computer scientists and the students largely want high paying jobs, not to continue in academia pushing the bounds of theory.
- UK-Al05 6y agoMost degrees in most subjects go in-depth on subjects they probably won't ever use in bog-standard industry role. Do I regret it? No, it was fun, interesting and it's cool when this hidden knowledge does come up at work.
- kodt 6y agoTrue, but it seems traditionally many schools have offered Computer Science, but not any other type of Software Engineering or programming degree and CS became the de-facto "I want to be a programmer" degree.
- stinkytaco 6y agoI've hired people, but wouldn't say I have a great deal of experiencing in the area, so here's an honest question for you: do candidates with a college education perform better at their jobs? In my (limited) experience the answer is definitely yes. Not only do they have more exposure to the concepts we need them to understand (if not necessarily the specific tools), we have just had better luck with them as employees in general. They are more likely to do small things, like all the administrative overhead that's part of a job: responding to requests in a timely manner, having good written communication skills, showing up to meetings on time, etc. I am the first to admit this is a small sample size and a lot of the cultural fit could be that the workplace is already filled with college educated staff, but I'm curious what your experience is after the hiring process is over.
- leoedin 6y agoThis may be the case, but it doesn't mean that those skills came from college. There's a strong selection bias at work - people with good administrative skills and communication skills are much more likely to go to college in the first place. We need to be asking if those people needed to go to college to develop those skills. In my experience at university, mostly the people who were the strongest - the people you wanted on your teams - were strong right from the start. They had some inherent skills which university developed and honed, but didn't create.
- stinkytaco 6y ago> There's a strong selection bias at work - people with good administrative skills and communication skills are much more likely to go to college in the first place. We need to be asking if those people needed to go to college to develop those skills. Agreed, but a university education reveals those people to me in a way that other evaluation criteria I have will not. I have lots of tools to evaluate whether a person has basic skills, but relatively fewer to determine whether they can do that other stuff. This is especially true as the reference and employment verification process is more limited than it used to be. Most companies I call can't tell me "don't hire this person" like they did 15 years ago. It doesn't help that, working for the government, I have limited options once I do hire someone. I am under pressure to make the right decision the first time since unmaking that decision is a great deal of painful time and effort. I suspect a number of mid-level managers like myself are in that position, which is why the degree continues to carry so much weight.
- bsurmanski 6y agoFor me, the biggest benefit of the university structure was 1) teaching me things I didn't even know existed and 2) teaching me things I didn't think we're important (but turned out to be). (1) might be solvable with an online curriculum, but (2) requires an evaluation or end-game credential to force me to do it. Other than that, as you mentioned an even bigger benefit was social, the shared experiences with others, and the cross pollination of ideas from these other people
- xvedejas 6y agoIt's not necessary to attend university to learn things you didn't know existed, or to learn things that seem less important in the moment. Whenever I've taught myself something out of interest, when I've seriously set aside time to explore a topic, it's not hard at all to expose oneself to many new concepts. This is assuming the field has a wealth of literature to read, and CS is one of those fields. Whether you have access to the internet or hardcover books, it'd be difficult not to stumble upon writing that is at least as compelling as the average college professor's lectures. Disclaimer: I am self-taught in CS, but also have an Engineering degree. I entirely agree that the social aspect of university is the largest benefit.
- carlinmack 6y agomaybe so, but there is a certain level of self-discipline and privilege needed to be able to study just from books. Lots of people on CS and other courses want friendly exposure to new ideas, not a recommendation of a list of text books
- _curious_ 6y agoDo you view the ability to study and learn from books as a privilege?
- dougmwne 6y agoStandard teacher education says that each student has a learning style that is most effective for them. https://www.learning-styles-online.com/overview/ https://www.learning-styles-online.com/overview/ A person who learns well from books in a solitary setting is at a huge advantage compared to someone who needs a social-kinesthetic style.
- mikece 6y agoOne of the benefits of a college setting would be the gathering together of young intelligent people and exposing them not only to new ideas but how to evaluate, value, and compare ideas for their relative worth. I wonder how much open and free debate of ideas even exists anymore, having given way to the inculcation and indoctrination of "correct ideas." I have long believed that the best way to eradicate racism is to allow racists and anti-racists to have open Lincoln-Douglas style debates on college campuses: allow racists to explain their ideas and positions and the tenability of their position will be exposed for the naked fear and hatred that it is. While not quite as crucial as the previous idea, debates of ideas, languages, operating systems, etc can and should happen as part of a CS curriculum. Regardless of the field we end up in, the need to be able to state one's position and defend it against a counter-proposal is something common to all fields (not just law).
- chadlavi 6y ago> allow racists to explain their ideas and positions and the tenability of their position will be exposed for the naked fear and hatred that it is People are pretty susceptible to passionate speech by charismatic figures; much more susceptible, in fact, than they are to facts. This just seems like an opportunity for racists to win hearts and minds. And anyway, there's tons of open and free debate on college campuses (at least, there was, when they existed in The Before Times), but generally it's debate about methodologies, not ideas. Everyone's more or less on the right (which is to say, left) side of history, they just have different ideas about, how to, for example, dismantle systemic racism, or end homelessness. Though probably the more frequent topics for debate are more along the lines of beer pong strategy.
- bcgraham 6y ago> People are pretty susceptible to passionate speech by charismatic figures; much more susceptible, in fact, than they are to facts. This just seems like an opportunity for racists to win hearts and minds. The unstated assumption of this premise is that some people (never us, always others) cannot be trusted to make good decisions, and thus need to have their decisions made for them, for their own good. It advocates for human subjugation.
- 0d9eooo 6y agoAs a former tenured professor I've been watching all of this with a bit of empathy, anxiety, concern, and schaudenfreude. Many of my best friends and colleagues are at the university and it still forms a huge part of my life. In speaking with many students and colleagues, one thing has struck me, which is that many of the current undergrads are enrolled because they want to get away from home. They want some true transition to adulthood, where they live on their own and develop into an independent adult with a vocation and adult identity. This is maybe what I've heard articulated here and elsewhere about being exposed to new ideas, etc. but maybe isn't quite the same. It's something I never really quite appreciated the importance of until recently, but looms large in their minds. Many of them, even though they're taking courses "off-campus" and online, are still in their college town living in their apartment, and so forth. They didn't go home, they just stayed put. The analogy with books and music is telling because for awhile there too there was a lot of hype about how the internet would lead to some radical reorganization and disruption of those industries. And in many ways it did. But many of the power inequities, the monopolies, and so forth are the same, and the fundamentals are the same. We've migrated a lot from radio and physical media to streaming services, but the players are still there. I suspect something similar will happen with universities; we'll just see a lot more diversity in what form an undergraduate education actually looks like. For some it will involve 4 years in a dorm / on-campus setting. For others it might be more eclectic. There are also some things that just can't really be duplicated at home: an undergrad isn't going to go out and purchase a mass spectrometer for the most part, or PCR equipment, or have storage room full of embalmed anatomy specimens to dissect. One thing that's important to be mindful of is that this on-campus experience many of us remember is in many ways evaporating anyway. I don't mean to say that's ok, or acceptable, or whatever, but sometimes we talk about the experience of being on campus and discussing things with others in person, forgetting that a major societal issue right now is how that that used to be the norm for everything and is increasingly not. So we have these chats online, on HN comment sections, or over IM, or videoconferencing anyway, even if you're not in college. Whatever might be lost from the on-campus experience is being lost everywhere to some extent; I point this out only to suggest that if we're ok with it in some domains, why should college be special? Or is college the canary that we should be paying closer attention to? I agree with the perspective of the author of the article that something fundamental needs to change, but my guess is 10 years from now everything will look very different and the same at the same time. I also think there's many problems with universities and higher education that have nothing to do with the cost-benefit calculus in some abstracted sense, but rather to do with how it should be paid for, whether universities are funded well enough, whether they're getting those funds from the right places, whether administration is too top-heavy, whether employers and society abuse interpretation of degrees, and so forth and so on.
- burntoutfire 6y agoHere in Poland there's a split between Computer Science (mostly lots of theory and fundamentals) and Software Engineering, which gives you some of those as well, but also prepares you for a typical programming job in a typical company.
- senux 6y agoI'm pretty certain that's true for a lot of countries other than the US. It makes sense after all. Far from American Universities to make any changes to their curriculum to support people's actual objectives though.
- jeltz 6y agoSame here in Sweden, and I think also the rest of the Nordic countries.
- viktor4006094 6y agoIt's the same here in Sweden. We have degrees called Högskoleingenjör (which is three years) and Civilingenjör (which is five years) that put more focus on practical and industry-relevant courses than the more science-focused Bachelor of Science and Master of Science programs do.
- WrtCdEvrydy 6y agoHonestly, the CS degree is weird. Some classes are super basic and some classes are nintendo hard. The classes that are in the middle of the road are very few but those are the ones that are most useful. Operating Systems, Data Structures, Parallel Computing and Software Engineering II (Basic Project Management) were notable for being interesting, difficult enough to make not slack and endlessly applicable for programmers. The CS degree isn't for programmers as my old professor once pointed out in a high level class: "Computer Science is about Programming like Biology is about pretty girls and microscopes". Those looking to go into research can take the CS coursework and probably get more use out of it.
- acemarke 6y agoI've always felt that a CS degree (at least based on the curriculum I went through) is really three separate and only partly related areas of study crammed together: - Teaching people how to program for the first time. Most of us had never actually programmed before, so you've got to have the intro classes that both teach _how_ to think through breaking a problem down into steps, and also the specific syntax of expressing those steps in a specific programming language, moving on to the typical concepts like OOP and classes and whatnot. - Actual CS. Data Structures, Algorithms, Operating Systems, Compilers, and the various more math-ish / "fundamentals" content. - Software Engineering. The practical, hands-on work of actually writing meaningful code, working as part of a team, requirements, using version control, testing, etc. Some classes like specific web dev topics or something probably fit into here too. Granted, I was going through this almost 20 years ago, and I only have my own experience to base this on, but it seems like this is a reasonable broad description of what a CS degree is trying to cover. I feel like that gave me a good fundamental background, but at the same time 99% of what I do on a daily basis now is all stuff that I either learned on the job, or in my own free time.
- WrtCdEvrydy 6y ago> I was going through this almost 20 years ago 10 here.... your comment holds up.
- graycat 6y ago> ... was a healthy debate about whether the typical CS curriculum helps a modern software engineer at all. Some thought the degree was stuck back in the 70's. Others thought there was value, but had trouble articulating it. I'll try a resolution of this dilemma: A few years ago I was talking with a university CS (computer science) prof, and his fast reaction was, IIRC, "The purpose of computer science is to identify the fundamentals of computing." I.e., make a scientific study of computing and find its basic laws. E.g., there is the subject of computational time complexity that for a given operation tries to find the fastest possible algorithm. Turns out, this is not an easy subject, e.g., we are still struggling with the question of P versus NP; a lot of well informed people believe that solving that problem would be a biggie. My guess is that the issue is cultural in that Einstein's E = mc^2 was seen as fundamental to the universe and other fields of research would like to get results also at that level of importance, fame, prestige, whatever. So, apparently in computer science making a scientific or engineering study of software engineering mostly loses out to pursuing some version of E = mc^2 for computing. E.g., one such result was the Gleason bound (in the Knuth volume on sorting and searching in his The Art of Computer Programming) that shows that O(n log(n)) is the fastest possible worst case sorting of n items by comparing items (records, keys, ...) two at a time. One response is that if medical schools were run like research university computer science departments, then no one would want to go to a hospital no matter how badly they hurt! I.e., a researcher deep into P versus NP might have never had occasion to use containers, Lisp, RUST, or even SQL. Of course, in medicine we have some medical schools in university research teaching hospitals where some of the best medical research is done. And it is easy for a practicing physician to conclude that mere clinical experience is not enough for the needed progress and that fundamental research in biochemistry, cell biology, DNA/RNA, the immune system, etc. is necessary, crucial, and really the best hope for the needed progress. So, it is fair to say that software engineering could be taught like medicine, not just as introduction to research but for practice, clinical practice, professional practice. E.g., if want to be the CIO of a large bank, insurance company, shipping firm, on-line retailer, etc., then how to manage an information technology (IT) department with 2000 employees to do well on data security, reliability, performance, software quality, software documentation, software system designs and revisions, growth, system management, etc.
- mattacular 6y ago> employers often look for the degree, but do not believe it is much of an indicator of quality and don't expect new grads to walk in the door with the needed skills In other words, requiring a college degree (even in an irrelevant field) of applicants is a form of class gatekeeping for technical professions.
- _curious_ 6y agoFeels like that sometimes, doesn't it? Then you find an employer in said technical profession who understands how their field has become inefficient at sourcing new employees and they seize on the opportunity to open the gate and let 'em in, thus creating a competitive advantage in the hiring pipeline and diversity in culture. The real tragedy is when all the others who can't/won't/don't open the gate cry whine and complain about not being able to find talent. Sometimes, when they are really pathetic, they have the gall to go to their local economic development agent with their hands out for a subsidy for hiring assistance. So those who adapt to the times end up paying taxes into a system that, in turn, gives that money (in part) those who don't. Perverse much?
- rs999gti 6y ago> indicator of quality Any other hiring managers have difficulties with Bachelor of Arts in CS versus Bachelor of Science? The BA applicants don't seem to have the logic and concept knowledge that the BS applicants have.
- Spooky23 6y agoMany of these arguments boil down to "those damn lazy kids these day". Being in a learning focused environment has value. I have a dual CS/History bachelor's degree. The analysis/rhetoric and writing skills that I built with the History side of my education are at least as valuable to me as the CS skillsets. I'd go back and challenge original implication of the article, that university is absurd, like a CD. Characterizing a CD as absurd is pretty short sighted -- it's the standard buy vs. rent argument. If you've ever followed a sport that has been in the middle of a cable contract dispute, the downside of the stream is obvious -- broadcasting is awesome until it isn't. I still buy music CDs or iTunes tracks. Access to an unlimited catalog doesn't do much for me, and owning perpetual rights is, to me, a better value, as it is for many things. That doesn't mean that streaming is "bad", just as online learning isn't "bad". But universities aren't obsolete because other options exist.
- kevin_thibedeau 6y agoI just ripped a new CD yesterday. I get a holistic album experience with all pregaps preserved. Something streaming and digital downloads won't provide.
- jimhefferon 6y agoWhen Netflix appeared, more movie watching happend. When MP3's appeared, more music got listened to. But when my school broke for COVID, the amount that got learned dropped sharply. I teach at a college but my wife teaches at a high school and it was the same there also. OK, maybe instructors are not up on what to do; for sure, I'm not. Maybe it is that the tech is not there yet; it all seems to me to be so static. But perhaps also the great majority of 18, 20, and 22 year olds aren't going to learn the same amount of material if they are at home also working at Target than if they are on campus with someone teaching them. I'll venture that certainly people who have significant socioeconomic and cognitive struggles are going to do worse in a climate where they are on their own. So I worry a lot about what would replace in-person college. The top post about how a person can make and do wonderful things is ... wonderful. But very, very few people will do it. Not disparaging that poster at all, in fact quite the opposite, just making a historical observation. > So that leaves the on campus experience, which I personally think is amazing. ... The new ideas from the classes are just a small part of the new ideas you get exposed to from other students and campus events. Plus with the increasing independence, it's a great gentle introduction to adulthood. Yes, it is great to be given space to grow. > employers often look for the degree Yes, having access to a good job is an important part of what people need to get. Of course, there are other things. Some understanding of civic responsibility comes to mind in the current climate, for instance. Even if we just focus on work, say with a CS degree, evidence to employers that candidates passed being expected to write well in a political science class, to understand something about data in a statistics class, and to have an overview of their field of speciality in a theory class, are things that they may be glad they have in an employee a decade from now. I get that many employers may be looking to what they need now, but I'm only saying that it is not dopey for at least some of them to be thinking along those lines.
- ardy42 6y ago> As the linked article points out, employers often look for the degree, but do not believe it is much of an indicator of quality and don't expect new grads to walk in the door with the needed skills. As a former recruiter I think that sentiment is pretty spot on. I don't think I ever asked a candidate about the details of their education. It was just never relevant in comparison to their experience. Universities aren't tech schools; their purpose is not job training. My understanding is that employers used to be comfortable providing pretty extensive job training, and they probably should start doing that again. IMHO, it's wishful thinking to expect any kind of school to produce new grads that can "walk in the door with the needed skills," especially in a field as diverse as software engineering.
- ntsplnkv2 6y agoCS is more applicable to mathematics or Physics. What CS needs is an equivalent engineering program. It is amazing to me that Software Engineering is not considered a valid engineering discipline like Civil, Mechanical (which apply maths and physics).