Is going to university worth it?

[teo123]

So, what do you guys on this forum think, is going to university worth it? Should somebody who wants to be a programmer go to university? I am not talking about programming nuclear reactors or medical devices, obviously, people doing that need some sort of qualification, because there is a very high risk. I am talking about the things most programmers do, such as, well, basically anything other than that. There is little risk if a script on a web-page malfunctions and I think that going to university for stuff like that is worse than useless.
In my opinion, and I think it’s an informed opinion, many of the stuff that we are being taught at the university are not only not useful, but are worse than useless. In our operating systems classes, we learn details about system calls, and how to do parallel programming by using system calls. Well, one thing I think nearly all programmers would agree on is that using system calls is a big no-no in programming. You should use standard libraries or other portable libraries, which solve the problem a lot better and in a way that won’t create countless problems once you try to run your program on some other computer architecture. And the stuff we learn in object-oriented development… Look, the worst programmers are definitely those who say Well, we’ve run into a problem because we didn’t obey the academic rules. Had we obeyed those academic rules, we wouldn’t have those problems. That’s probably both false and insulting. Yet, it seems to me studying object-oriented development does exactly that. I think software architecture is, at best, as useful to programmers as ornithology is to birds, yet that universities insist we know that in details. I also fail to see how it can be called science. They don’t seem to cite any statistics, yet alone something more.
I also have some idea that I think I could be making money with, but, in order to realize it, I need to dedicate a lot of my time to that. Namely, I think I’ve thought of a language that would be a good alternative to AssemblyScript. I’ve been writing a compiler and, after I’ve written some 3’000 lines of code, I’ve managed to compile a simple program written in that language so that it can run in a browser. That looks relatively promising to me, however, as I’ve failed 5 exams this year, I probably need to choose either going to university or continuing to develop that program. So, what do you think, is it a good idea to drop out from the university?

[NonZeroSum]

I also have some idea that I think I could be making money with, but, in order to realize it, I need to dedicate a lot of my time to that. Namely, I think I’ve thought of a language that would be a good alternative to AssemblyScript. . .
So, what do you think, is it a good idea to drop out from the university?

No, don't trust your instincts to know the probability of what a good business venture is, you will just be holding a ton of assumptions that you've convinced yourself into just like flat earth.

Learn to appreciate the skill of mastering a task fast even if you'll never use it, because learning is the skill in and of itself, like learning to speed read so you can take in information faster.

If you drop out you will likely be taking worse jobs that rack up more hours of the mundane experience you feel taking classes you don't like now, than if you just worked hard and got a 1st class degree to go on your job portfolio.

[teo123]

No, don't trust your instincts to know the probability of what a good business venture is, you will just be holding a ton of assumptions that you've convinced yourself into just like flat earth.

Sorry, I fail to see the connection between this and the Flat Earth Theory. This works, as you can see for yourself. It's not yet finished, but, unlike most of the stuff we learn at the university (and unlike the Flat Earth Theory), you can actually see that it works.

Learn to appreciate the skill of mastering a task fast even if you'll never use it, because learning is the skill in and of itself, like learning to speed read so you can take in information faster.

To me, saying that sounds like wishful thinking. Look, if you want to develop a website using some new technology (new JavaScript framework, new PHP framework...), would you ask for help somebody who already knows a few JavaScript and PHP frameworks, or would you ask somebody who knows a bunch of unrelated stuff?

If you drop out you will likely be taking worse jobs that rack up more hours of the mundane experience you feel taking classes you don't like now, than if you just worked hard and got a 1st class degree to go on your job portfolio.

Hey, listen, there are studies suggesting that, but, as Brian Caplan says in "The Case against Education", almost all of those studies are fundamentally flawed. People who go to university are people who did better in high-school. It's to be expected that people who did better in high-school tend to earn more than people who did bad at high-school, regardless of the university. And even if the university does have some effect, is it enough to justify spending five years studying difficult things you don't like?

[brimstoneSalad]

There are some people for whom university is not useful, but like @NonZeroSum said you are not one of those people. You would very much benefit from the structure and learning how to learn aspect of university because you tend to go off in very nutty directions when left to your own educational devices. Do not trust your business intuitions.

[Red]

It's important to note that many business ventures end in failure because the people overestimated their ability or underestimated the weight of certain risks.
From a financial standpoint starting a business is just fine as long as you have something to fall back on (usually a university education in STEM or something) since those taking off are not very likely. Like with me, I hope my YouTube channel can get off the ground so I can make some money off of that and patreon, but I'm also going to Uni for a degree in engineering in case that doesn't work out.

[teo123]

learning how to learn aspect of university

But, you know, there is a vast difference between learning to pass a test and learning in order to be able to build something. Techniques of learning that work in the former case are counter-productive in the latter case. If you are learning in order to build something, your goal is to distinguish what will be important to you from what will not be important to you. When learning to pass some test, you are trying to guess what the professor thinks is important (for whatever reason).

from the structure

I am not sure what you mean, can you clarify?

you tend to go off in very nutty directions when left to your own educational devices

Well, that really can't happen in computer science, because I can see whether the things I made work. You can see it too, now it's runnable in a browser. Sure, there is a bit of soft science regarding guessing whether the language I've designed is actually good for the purpose of competing with AssemblyScript, but that's not remotely as soft as guessing, for example, what's a good user interface.

Do not trust your business intuitions.

Well, even if I fail, I'll still learn a lot of useful things (a lot more than at the university) which will help me succeed next time, right?

I hope my YouTube channel can get off the ground so I can make some money off of that and patreon

Well, that's different. There is no actual need for a new YouTube channel and you can only guess whether people will like your content (which is entirely subjective and very hard, if not impossible, to predict). Furthermore, almost no YouTubers make money on YouTube. On the other hand, there is obviously a need for languages with nicer syntax to be run on WebAssembly. All languages which are currently available are C-like languages, which many people consider a bad thing, and for obvious reasons. Furthermore, it's not at all obvious there are many people who have tried to make money by making a programming language and have failed very badly.

[Red]

Well, that's different. There is no actual need for a new YouTube channel

I highly doubt a YouTube channel promoting science and veganism isn't more important than another programming language.

Furthermore, almost no YouTubers make money on YouTube.

Where are you getting this from? If you took every channel into account sure, but 90% of YouTube channels don't upload anything. Anyone moderately successful on YouTube can make decent money from ads and Patreon.

On the other hand, there is obviously a need for languages with nicer syntax to be run on WebAssembly. All languages which are currently available are C-like languages, which many people consider a bad thing, and for obvious reasons. Furthermore, it's not at all obvious there are many people who have tried to make money by making a programming language and have failed very badly.

What makes you different?

[NonZeroSum]

You know what would be a great therepeutic tool for making sure you have a good judge on the probabilites of what's the best career path to take. Confronting your past probablistic mistakes. Walk down to the local jail at visiting hours. Just say the word that you'll do it with a video camera and I'll even set up a go fund me account and promote the shit out of that thing, so you'll have the knowledge you're funding a good cause too.

[teo123]

I highly doubt a YouTube channel promoting science and veganism isn't more important than another programming language.

First, channels talking about veganism (positively or negatively) tend to get demonetized and hated. I've published a few videos about veganism/vegetarianism, only the video about vegetarianism in Latin has more likes than dislikes (and the comments are rather negative). I wouldn't expect to be making money out of promoting veganism. That's not to say it's unethical to do that.
Second, there, as far as I know, just aren't many programming languages that can be used to target JavaScript Virtual Machine, or WebAssembly. Out of primary languages used to target it, JavaScript, AssemblyScript, TypeScript and CoffeeScript, only CoffeeScript doesn't have C-like syntax. And CoffeeScript is in many cases unacceptably slow, because it transpiles to JavaScript, which is itself slow (Because JavaScript wasn't at all designed with speed in mind. Or, for that sake, anything other than making any kind of language that can be run in a primitive browser as soon as possible in mind. It was made in 10 days.). Rust is also sometimes used to target JavaScript Virtual Machine, and it also has a C-like syntax, and a weird semantics which, while I can see how some people might like it, it's not for everyone. I get very annoyed when the compiler refuses to compile my code just because people who made the language consider something a code-smell, especially when there is no obvious way to fix it. And, using Emscripten, you can use C and C++ to target JavaScript Virtual Machine, which are again C-like languages. Furthermore, scripting an actual web-page using C++ is next to impossible because of the difficulty of passing strings between C++ and JavaScript. When writing a program to be run in a browser, the number of languages you can use is very limited. You can't even use Ada, which can even be used for writing kernels.

If you took every channel into account sure, but 90% of YouTube channels don't upload anything.

Then which metric would you suggest to estimate the probability of somebody succeeding as a YouTuber?

Anyone moderately successful on YouTube can make decent money from ads and Patreon.

That's like saying anyone moderately successful in singing can make money of singing or that anyone moderately successful in acting can make money as an actor.

[Red]

First, channels talking about veganism (positively or negatively) tend to get demonetized and hated.

Any channel that talks about a controversial subject will have its detractors. This doesn't just apply to vegan YouTube channels. Talks and documentaries are also met with negative reception from certain groups of people. How do you suggest people advocate for veganism then?

It isn't like making a YouTube channel promoting veganism and science won't influence anybody. It's just a matter of communicating the subject properly.

I've published a few videos about veganism/vegetarianism, only the video about vegetarianism in Latin has more likes than dislikes (and the comments are rather negative).

Well yeah because your videos are terrible. People want to watch better edited and well scripted videos.

I wouldn't expect to be making money out of promoting veganism. That's not to say it's unethical to do that.

Veganism isn't the only thing I'm gonna be talking about FYI.

Second, there, as far as I know, just aren't many programming languages that can be used to target JavaScript Virtual Machine, or WebAssembly. Out of primary languages used to target it, JavaScript, AssemblyScript, TypeScript and CoffeeScript, only CoffeeScript doesn't have C-like syntax. And CoffeeScript is in many cases unacceptably slow, because it transpiles to JavaScript, which is itself slow (Because JavaScript wasn't at all designed with speed in mind. Or, for that sake, anything other than making any kind of language that can be run in a primitive browser as soon as possible in mind. It was made in 10 days.). Rust is also sometimes used to target JavaScript Virtual Machine, and it also has a C-like syntax, and a weird semantics which, while I can see how some people might like it, it's not for everyone. I get very annoyed when the compiler refuses to compile my code just because people who made the language consider something a code-smell, especially when there is no obvious way to fix it. And, using Emscripten, you can use C and C++ to target JavaScript Virtual Machine, which are again C-like languages. Furthermore, scripting an actual web-page using C++ is next to impossible because of the difficulty of passing strings between C++ and JavaScript. When writing a program to be run in a browser, the number of languages you can use is very limited. You can't even use Ada, which can even be used for writing kernels.

And what are the chances that you will be able to do it? If there isn't much attention to things like this it probably isn't very significant. More qualified people can handle this I think.

Then which metric would you suggest to estimate the probability of somebody succeeding as a YouTuber?

The most successful YouTube channels tend to make videos that people want to watch. Of course, algorithms play a part, but if you upload decent videos it gives you a chance.

That's like saying anyone moderately successful in singing can make money of singing or that anyone moderately successful in acting can make money as an actor.

Yeah that seems about right. Your point?