Is learning to code a waste of time?

[Alchemisst]

I just watched the title youtube video, and it kind of put me off programming, kind of makes it seem like its a waste of time etc. and not really useful and what you learn will be redundant in a few years (you'll never become an expert over time like other careers such as law, accounting medicine etc) it basically seems to say don't learn to code but doesn't really give an alternative?

[clarkfan1979]

If you are a good coder, you get paid well. My own personal bias is that coders tend to be one dimensional. Because they have the option to "work from home" they tend to be working all the time. They don't really have time for anything else and hire most projects out. Many of them have good salaries, but also high expenses and pretty low net worth. This is just my bias from my social circle.

[dividendman]

Coding + computer science = 150k starting comp minimum if you're any good.

[reeshau]

Learning any particular programming language is a skill that could become obsolete, but learning to code, that is good program design, testing / QA, optimization and refactoring, etc. , is something you will always use.  Picking up a new language is either an easy, enjoyable opportunity or just a maintenance activity, depending on your view.  You simply need to have a first language to learn underlying principles on; if you're paying attention or in a good educational program, you will pick this up.  If you go to any number of computer science mills, you will just learn a language.

I would add that there are plenty of programs out there that have lived long after their creators thought they would be obsolete.  There is currently a large number of open Cobol positions just for this reason: It's viewed as obsolete, so no new entrant wants to learn it.  But as its practitioners retire, the skill has become precious.

[fell-like-rain]

Like many things, the answer is “it depends”. Do you have something specific you want to do, whether that’s make a video game or a phone app or a script that texts your teenager “cut it out!” every time they raise the temp on your smart thermostat? Or are you looking for a career path in software? If the answer to both of those is no, it’s a useless skill.

[maizefolk]

Learning to program as a career has its pluses and minuses. You can make a lot of money right out of school, but you do have to always be watching out to make sure the skills on your resume pass their sell by date. It's also a field with a lot of ageism against the those in middle age or beyond, so you probably don't still want to be writing code in your 40s or 50s unless you've found a secure long term job by then.

Learning to code as an add on skill to make you faster and more productive in any other career path (automate the boring parts of your job!), now that has nothing but pluses. For this purpose the language doesn't particularly matter, nor does writing particularly good or clean or maintainable code. I have colleagues who only know perl, aren't actually that good at it, and do things like copy and paste blocks of code rather than write a function. They are still more productive, and spend less of their life doing boring repetitive work, than my colleagues who cannot program at all.

[Dave1442397]

Learning to code is a small part of being a good coder, in my experience. You want to be the coder that management requests for the projects with drop-dead due dates, or the complicated stuff that requires logical thinking and questioning the Business Analyst when things don't look right.

My employer insisted that we use offshore coders for a few years. They were all straight out of college, so they knew (more or less) how to code, but they only did what they were told, even if it was obviously wrong. One funny example: A Business Analyst sent in a screen redesign, but they had done it in Word and didn't format it properly, so the offshore programmer did exactly what they saw, which was a screen with mis-aligned columns, colons in different positions on every line, etc.

We finally got rid of them when upper management figured out that all the extra time being spent on projects was because we had to fix everything they did, usually at least twice, before we could get the code out to the clients.

Everyone who codes in my department is over 50. One is 71 (don't you have a life?). They are all really good at what they do, and the application we work on processes a large percentage of all the daily mutual fund transactions in the US every night. We had a few days in March that went over three trillion dollars worth of transactions, so you don't want to be screwing up the code.

[Tester]

Learning to code is different from learning how to build something.
It was already touched on.
Learn systems theory, logic, then look into coding.
If you want to learn to code for the money you might find out it is not that enjoyable.
It is still a good think to pick up, but only if you first learn to think like an engineer.

[MilesTeg]

Coding + computer science = 150k starting comp minimum if you're any good.

LMFAO not even remotely the case, except perhaps in extremely HCOL areas.

[MilesTeg]

I just watched the title youtube video, and it kind of put me off programming, kind of makes it seem like its a waste of time etc. and not really useful and what you learn will be redundant in a few years (you'll never become an expert over time like other careers such as law, accounting medicine etc) it basically seems to say don't learn to code but doesn't really give an alternative?

Not a waste of time if you focus on on concepts (problem solving, machine/network architecture, language concepts, etc) rather than being a tool monkey. It's very true that any particular tool you might learn will become obsolete very quickly, but that doesn't mean core skills will.

In my 20 year career I have done significant work in about two dozen major languages and likely hundreds of different tools. Am I an 'expert' in all those languages? Hell no. What I am an expert at is the core concepts of software engineering that I can apply with any tool I am asked to use. Part of being a good software engineer is developing the ability to swap between tools with little friction.

[SEdude]

Coding + computer science = 150k starting comp minimum if you're any good.

To anyone reading this thread considering a career in software, I just want to point out that the above statement is hyperbole. Outside of a handful of big software companies (facebook, amazon, apple, netflix, google and a few of the 'unicorn' startups), all of which are in very high cost of living areas, it is very unlikely to have entry level compensation of 150k. And as you would expect of such sought after jobs, entry level positions at said companies are highly competitive (and are biased towards ivy league / stanford / mit graduates). So plenty of people that are "any good" don't start at 150k.

That said, you can definitely make a lot of money in software, and 150k is an attainable compensation for most people in the industry after several years if you live in SF, NYC or seattle.

[zolotiyeruki]

I just watched the title youtube video, and it kind of put me off programming, kind of makes it seem like its a waste of time etc. and not really useful and what you learn will be redundant in a few years (you'll never become an expert over time like other careers such as law, accounting medicine etc) it basically seems to say don't learn to code but doesn't really give an alternative?

Not a waste of time if you focus on on concepts (problem solving, machine/network architecture, language concepts, etc) rather than being a tool monkey. It's very true that any particular tool you might learn will become obsolete very quickly, but that doesn't mean core skills will.

In my 20 year career I have done significant work in about two dozen major languages and likely hundreds of different tools. Am I an 'expert' in all those languages? Hell no. What I am an expert at is the core concepts of software engineering that I can apply with any tool I am asked to use. Part of being a good software engineer is developing the ability to swap between tools with little friction.

^^^ This.  I've done a fair bit of coding in probably half a dozen languages, and work with a bunch of other developers.  What I've seen is that problem solving ability, adaptability, work ethic, independence, etc are just as important as, if not more important than actual coding ability.

When it comes to coding, there's always some new fad around the corner, some new language that someone has invented to resolve imperfections (or, often as not, perceived imperfections) in existing languages.  Apple basically reinvents Swift with each release (oh, how I pity iOS developers).  Depending on where you work, you may or may not be subject to the whims of the industry.  Having a solid foundation in C/C++/Java will set you up nicely for learning just about any other language, and when the Framework Of The Day goes out of fashion, you'll still have those foundational skills.

[BlueMR2]

Reflecting on nearly 30 years of coding things actually used by other people...  If you enjoy it, go for it.  If not, then don't.  Coding is the kind of thing that requires continual reinvestment in yourself to stay current.  Odds of you doing that if you don't enjoy it are slim.

It's also filled with people making 35K a year.  There are certainly people making lots more, but those are the exceptions rather than the rule.  Skill doesn't get you those high paying jobs either.  Those are basically luck of the draw.

[PaulMaxime]

I just watched the title youtube video, and it kind of put me off programming, kind of makes it seem like its a waste of time etc. and not really useful and what you learn will be redundant in a few years (you'll never become an expert over time like other careers such as law, accounting medicine etc) it basically seems to say don't learn to code but doesn't really give an alternative?

Here's the thing. i've been writing code professionally for over 30 years. I've never felt any risk of being redundant or out of date. Not once. The key is that I understand the fundamentals of computing which never change. Sure you have to learn new concepts and languages over time but that's actually super easy. These days the tools are good and they actually understand the syntax of languages and you also get to look anything up on the internet any time which is a big advantage over back in the day when you had nothing but a plain text editor and a book.

But anyone who tells you that you are going to become obsolete is just frankly not that good. In my career I've gone from building networking code to data analysis systems to designing security hacking tools and electronic book readers and email archiving systems to now building video games.

It's really all the same.

[gooki]

Reflecting on nearly 30 years of coding things actually used by other people...  If you enjoy it, go for it.  If not, then don't. 

This x1000.

[MilesTeg]

Reflecting on nearly 30 years of coding things actually used by other people...  If you enjoy it, go for it.  If not, then don't. 

This x1000.

Yep, the coding landscape is littered with the desiccated corpses of fools who try to do it because they heard you can make lots of money.

[ctuser1]

Lots of people provided very good insights.

I think the answer also depends on the background of the person asking the question, since that tells you how easy it likely will be to learn, and how useful it will be once learned.

Someone with a hard science background - well, picking up coding will not be too difficult for them. So it is a massive bang for the buck for them even if it is not of much use in their work. In fact, probably 99.99% of them would already be able to code precisely for this reason.

If you are an accountant - coding can be a major asset for your career because you can then automate away boring tasks, or effectively interface with the tech team. So, even if learning coding is more effort for this person compared to - say - a particle physicist, it will likely be a good value given how massive the payoff is.

If you are a high-school arts teacher - I doubt learning to code will either be easy or helpful in any way. So meh!

[ditheca]

... what you learn will be redundant in a few years (you'll never become an expert over time like other careers such as law, accounting medicine etc)

Why would a coding be any less demanding than law, accounting, or medicine? Every skilled profession benefits from continual education.

A lawyer who doesn't keep up-to-date with recent developments in the legal field is likely to get sued for malpractice.

My mother, (a tax accountant), spends roughly 5% of her time on studying and training/classes to be aware of changes to constantly changing tax laws.

A doctor who graduated 30 years ago and has be coasting ever since would not know about medical lasers, genetic testing, viagra, etc.

[PaulMaxime]

Someone with a hard science background - well, picking up coding will not be too difficult for them. So it is a massive bang for the buck for them even if it is not of much use in their work. In fact, probably 99.99% of them would already be able to code precisely for this reason.

Scientists write the worst code! It's so bad.

[maizefolk]

Someone with a hard science background - well, picking up coding will not be too difficult for them. So it is a massive bang for the buck for them even if it is not of much use in their work. In fact, probably 99.99% of them would already be able to code precisely for this reason.

Scientists write the worst code! It's so bad.

Am a scientist, can confirm this.

In fairness though a lot of scientist code is being written to run 1-3 times EVER. There's a lot of investment in readability/maintainability/scaleability/run time efficiency which pays off for code that'll be run thousands of times by hundreds of people and so is taught to people who write code professionally as something you always do but isn't actually worth the extra time writing the code when you can count total times being used on one hand.

.... but we're also just often bad at it.

[AccidentialMustache]

Learning to program as a career has its pluses and minuses. You can make a lot of money right out of school, but you do have to always be watching out to make sure the skills on your resume pass their sell by date. It's also a field with a lot of ageism against the those in middle age or beyond, so you probably don't still want to be writing code in your 40s or 50s unless you've found a secure long term job by then.

I'm beginning to think the ageism in tech is really just a coasts problem -- possibly just west coast -- possibly just silicon valley. I'm in the midwest and I see plenty of greyer beards than mine.

The sell by date thing may also be overrated. Let's talk about major languages you'd find on my resume over time, and their dates:

* Fortran (1954)
* C (1969-1973)
* SQL (1978)
* Tcl (1988)
* Python (1991)
* Ruby (1995)
* Java (1995)
* If current trends continue at work I might feel comfortable adding JavaScript (1995)/TypeScript (2012).

Of that set of languages, TypeScript is the only one released after I graduated high school. If you head to the start of the list you're looking at before I was born.

[BussoV6]

Learning to program as a career has its pluses and minuses. You can make a lot of money right out of school, but you do have to always be watching out to make sure the skills on your resume pass their sell by date. It's also a field with a lot of ageism against the those in middle age or beyond, so you probably don't still want to be writing code in your 40s or 50s unless you've found a secure long term job by then.

I'm beginning to think the ageism in tech is really just a coasts problem -- possibly just west coast -- possibly just silicon valley. I'm in the midwest and I see plenty of greyer beards than mine.

The sell by date thing may also be overrated. Let's talk about major languages you'd find on my resume over time, and their dates:

* Fortran (1954)
* C (1969-1973)
* SQL (1978)
* Tcl (1988)
* Python (1991)
* Ruby (1995)
* Java (1995)
* If current trends continue at work I might feel comfortable adding JavaScript (1995)/TypeScript (2012).

Of that set of languages, TypeScript is the only one released after I graduated high school. If you head to the start of the list you're looking at before I was born.

What, no Turbopascal or Delphi??  :-)

[Dave1442397]

I've been coding in Fortran since 2010. We've migrated to linux in that time period (last client moving from VMS this year), and all the surround systems have been modernized, but Fortran endures. Each new CIO says "Fortran? Why don't we convert that to <flavor of the month>?" Uh, we have 20,000,000 lines of Fortran code...where would you like to start?

[Gronnie]

Coding + computer science = 150k starting comp minimum if you're any good.

Only in Silicon Valley where your expenses will be super high as well.

[dividendman]

Coding + computer science = 150k starting comp minimum if you're any good.

Only in Silicon Valley where your expenses will be super high as well.

Nah, I helped open an Amazon tech center right there in Minneapolis and they paid just as high.

I don't know why people think it's a joke. Definitely new york, seattle, silicon valley you can get that much. But you don't have to take my word for it. Go on team blind or some other site where people share their comp and you can see people with starting compensation all over the US at over 100k.

[BDWW]

Scientists write the worst code! It's so bad.

Am a scientist, can confirm this.

In fairness though a lot of scientist code is being written to run 1-3 times EVER but will live on forever. There's a lot of investment in readability/maintainability/scaleability/run time efficiency which pays off for code that'll be run thousands of times by hundreds of people and so is taught to people who write code professionally as something you always do but isn't actually worth the extra time writing the code when you can count total times being used on one hand.

.... but we're also just often bad at it.

++We work constantly on research projects with universities, and I've half jokingly/seriously suggested we assign each professor/grad student a full time coder to follow them around and implement/fix their ideas so we don't have to deal with the rubbish they come up with.

[SEdude]

Coding + computer science = 150k starting comp minimum if you're any good.

Only in Silicon Valley where your expenses will be super high as well.

Nah, I helped open an Amazon tech center right there in Minneapolis and they paid just as high.

I don't know why people think it's a joke. Definitely new york, seattle, silicon valley you can get that much. But you don't have to take my word for it. Go on team blind or some other site where people share their comp and you can see people with starting compensation all over the US at over 100k.

I don't think it's a joke, I just don't think it is accurate for most fresh CS graduates. You are talking about the very top earning graduates (people who go on team blind are usually the ones who care about the "prestige" of which company to work for and getting the absolute highest compensation at google, amazon, facebook, etc), and it is very misleading to someone who is looking to get into the industry.

My anecdata from a top state school had 3 people north of 150k/yr starting (compensation, not salary. ie including stock grants), all in SF or seattle. A dozen or so in the 120-140k/yr range, in HCOL, and most under 100k in medium or low COL areas (probably 80k would be considered a good starting comp these days in a place like austin tx). I did my best to adjust the numbers based on today's levels, which I think are slightly higher than they were a few years back. My point is that yes, there are some people who get the compensation you're talking about, but it's the top of the top.

And the echo others in the thread, don't go into software just for the money. Only do it if you actually like it!

[Gronnie]

Coding + computer science = 150k starting comp minimum if you're any good.

Only in Silicon Valley where your expenses will be super high as well.

Nah, I helped open an Amazon tech center right there in Minneapolis and they paid just as high.

I don't know why people think it's a joke. Definitely new york, seattle, silicon valley you can get that much. But you don't have to take my word for it. Go on team blind or some other site where people share their comp and you can see people with starting compensation all over the US at over 100k.

I don't think it's a joke, I just don't think it is accurate for most fresh CS graduates. You are talking about the very top earning graduates (people who go on team blind are usually the ones who care about the "prestige" of which company to work for and getting the absolute highest compensation at google, amazon, facebook, etc), and it is very misleading to someone who is looking to get into the industry.

My anecdata from a top state school had 3 people north of 150k/yr starting (compensation, not salary. ie including stock grants), all in SF or seattle. A dozen or so in the 120-140k/yr range, in HCOL, and most under 100k in medium or low COL areas (probably 80k would be considered a good starting comp these days in a place like austin tx). I did my best to adjust the numbers based on today's levels, which I think are slightly higher than they were a few years back. My point is that yes, there are some people who get the compensation you're talking about, but it's the top of the top.

And the echo others in the thread, don't go into software just for the money. Only do it if you actually like it!

This is much more accurate.

[MilesTeg]

Scientists write the worst code! It's so bad.

Am a scientist, can confirm this.

In fairness though a lot of scientist code is being written to run 1-3 times EVER but will live on forever. There's a lot of investment in readability/maintainability/scaleability/run time efficiency which pays off for code that'll be run thousands of times by hundreds of people and so is taught to people who write code professionally as something you always do but isn't actually worth the extra time writing the code when you can count total times being used on one hand.

.... but we're also just often bad at it.

++We work constantly on research projects with universities, and I've half jokingly/seriously suggested we assign each professor/grad student a full time coder to follow them around and implement/fix their ideas so we don't have to deal with the rubbish they come up with.

I made my early career and a successful small business unit out of effectively doing that. However, I am glad I no longer have to make sense out of excel pivot table madness.

[ctuser1]

I made my early career and and successful small business unit out of effectively doing that. However, I am glad I no longer have to make sense out of excel pivot table madness.

I'll up that by one! Try array formulas!

They are not commonly used because they require you to think matrix math - and hence not easy to grasp. I had a PhD-in-Physics Quant dump onto me an excel madness full of array formulas that I had to make sense. I had to literally dust up my linear algebra textbooks, translate each opeartion into matrix operation notation, before I could get some sense of what was going on in that excel madness.

This lasted about 2 months!

Let's just say that is the only time in my life I have had suicidal thoughts - and I have dealt with many messed up projects. Indeed, taking messed up initiatives, figuring out how to straighten it out and then hand over to someone to actually do it *was* my speciality. But that sh*tty spreadsheet was just something else!!

[MilesTeg]

I made my early career and and successful small business unit out of effectively doing that. However, I am glad I no longer have to make sense out of excel pivot table madness.

I'll up that by one! Try array formulas!

They are not commonly used because they require you to think matrix math - and hence not easy to grasp. I had a PhD-in-Physics Quant dump onto me an excel madness full of array formulas that I had to make sense. I had to literally dust up my linear algebra textbooks, translate each opeartion into matrix operation notation, before I could get some sense of what was going on in that excel madness.

This lasted about 2 months!

Let's just say that is the only time in my life I have had suicidal thoughts - and I have dealt with many messed up projects. Indeed, taking messed up initiatives, figuring out how to straighten it out and then hand over to someone to actually do it *was* my speciality. But that sh*tty spreadsheet was just something else!!

I feel your pain, even though I have never had the displeasure of working on something like that.

[GuitarStv]

Coding + computer science = 150k starting comp minimum if you're any good.

Only in Silicon Valley where your expenses will be super high as well.

Nah, I helped open an Amazon tech center right there in Minneapolis and they paid just as high.

I don't know why people think it's a joke. Definitely new york, seattle, silicon valley you can get that much. But you don't have to take my word for it. Go on team blind or some other site where people share their comp and you can see people with starting compensation all over the US at over 100k.

I don't think it's a joke, I just don't think it is accurate for most fresh CS graduates. You are talking about the very top earning graduates (people who go on team blind are usually the ones who care about the "prestige" of which company to work for and getting the absolute highest compensation at google, amazon, facebook, etc), and it is very misleading to someone who is looking to get into the industry.

My anecdata from a top state school had 3 people north of 150k/yr starting (compensation, not salary. ie including stock grants), all in SF or seattle. A dozen or so in the 120-140k/yr range, in HCOL, and most under 100k in medium or low COL areas (probably 80k would be considered a good starting comp these days in a place like austin tx). I did my best to adjust the numbers based on today's levels, which I think are slightly higher than they were a few years back. My point is that yes, there are some people who get the compensation you're talking about, but it's the top of the top.

And the echo others in the thread, don't go into software just for the money. Only do it if you actually like it!

This is much more accurate.

An above average software developer in Canada with 15 - 20 years experience who is still actually doing coding is most likely making 80 - 150k per year.  Canadian.  :P

[FINate]

@Alchemisst Is learning to program a waste of time? Depends on what your goals are and how you determine success.

If the goal is a quick path to easy money then save yourself the time and expense. As the saying goes, there's no such thing as a free lunch. Shortcuts to programming careers result in lower paying coding jobs w/ reduced upward mobility. I personally know many people that took the quick path to a coding job that later regretted it. They all said essentially the same thing: It would have been less work and more lucrative to just get the CS degree from the start. Of course, there are exceptions. But I would caution anyone starting into a new field thinking they are exceptional ;-)

There's a cottage industry of experts in obscure but important technologies, including certain programming languages. These tend to be things like legacy enterprise technologies. Folks in these areas tend to have a lifetime of experience and they are few and far between. This is not something that one casually breaks into.

As for mainstream technologies: You would have to rise above the millions of people already proficient and experienced with these languages before being recognized as an expert. Typically this means being an early adopter of a new language that goes mainstream, being well known and trusted within that community, and then authoring books and giving lectures which substantially add to the knowledge and spread of said language. In other words, it's exactly the same as every other profession. Becoming an expert requires a lifetime of learning, intentional career growth, and commitment to promoting and contributing to the field.

Learning to code is a small part of a career in software and, IMO, one of the easier parts. Your question is analogous to asking if learning first-aid is a waste of time because, on its own, it doesn't provide a career path to becoming a paramedic.

The alternative is to focus on a long-term goal of a career in tech. A CS degree from a reputable institution is the most common starting point, but there are many ways to approach it.

[seattlecyclone]

Coding + computer science = 150k starting comp minimum if you're any good.

Only in Silicon Valley where your expenses will be super high as well.

Nah, I helped open an Amazon tech center right there in Minneapolis and they paid just as high.

I don't know why people think it's a joke. Definitely new york, seattle, silicon valley you can get that much. But you don't have to take my word for it. Go on team blind or some other site where people share their comp and you can see people with starting compensation all over the US at over 100k.

I don't think it's a joke, I just don't think it is accurate for most fresh CS graduates.

To be fair to @dividendman, they were referring to coders who are "any good." He didn't say anything about the population of CS graduates as a whole. My experience going through a mid-tier state school showed me that there are a rather large number of people who can acquire a CS degree without developing any particular aptitude for software development. I remember taking one class where we had to do pair projects and there was only one other guy in my class who I thought was "any good" (and he felt the same) so we became partners. That worked out great! (He ended up getting hired by Microsoft, btw). Then for the next round of projects they teamed each pair with another pair to make groups of four, and the pair we got stuck with was literally no help at all. Someone who kind of muddles through their algorithms and operating systems classes will earn a degree, but they might be "just OK" at software development, not "any good." The Big Tech firms try very, very hard not to get stuck with the "just OK" folks. The teams I was on before FIRE always had open positions, and not for lack of applicants.

[SwordGuy]

If you are a high-school arts teacher - I doubt learning to code will either be easy or helpful in any way. So meh!

And yet artists are now using 3D printing to make items.   It might be a finished piece of art or it might be used in mold-making to cast the final item.   To do so, they program using a graphic design script. 

Learning to think like a successful programmer is a very useful skill in ANY endeavor.    Programming doesn't teach logical thinking so much as it teaches the result of using muddled, illogical thinking, i.e., failure.   Most people actually have real trouble precisely and accurately thinking thru a problem to solve, people who learn to program learn how to do that much better.   Learning how to think thru and solve problems is a valuable skill in any language.

   

[dividendman]

An above average software developer in Canada with 15 - 20 years experience who is still actually doing coding is most likely making 80 - 150k per year.  Canadian.  :P

And that's why I left Canada. The same skills even adjusted for cost of living pay way more in the US.

[SEdude]

Coding + computer science = 150k starting comp minimum if you're any good.

Only in Silicon Valley where your expenses will be super high as well.

Nah, I helped open an Amazon tech center right there in Minneapolis and they paid just as high.

I don't know why people think it's a joke. Definitely new york, seattle, silicon valley you can get that much. But you don't have to take my word for it. Go on team blind or some other site where people share their comp and you can see people with starting compensation all over the US at over 100k.

I don't think it's a joke, I just don't think it is accurate for most fresh CS graduates.

To be fair to @dividendman, they were referring to coders who are "any good." He didn't say anything about the population of CS graduates as a whole. My experience going through a mid-tier state school showed me that there are a rather large number of people who can acquire a CS degree without developing any particular aptitude for software development. I remember taking one class where we had to do pair projects and there was only one other guy in my class who I thought was "any good" (and he felt the same) so we became partners. That worked out great! (He ended up getting hired by Microsoft, btw). Then for the next round of projects they teamed each pair with another pair to make groups of four, and the pair we got stuck with was literally no help at all. Someone who kind of muddles through their algorithms and operating systems classes will earn a degree, but they might be "just OK" at software development, not "any good." The Big Tech firms try very, very hard not to get stuck with the "just OK" folks. The teams I was on before FIRE always had open positions, and not for lack of applicants.

I think this is actually the crux of my issue with the original statement. If you are interpreting the phrase "any good" as "the top few people in class" then yes, I agree with dividendman that if you are one of those people you can earn 150k+ right out of college easily. However, I don't think that most people find the phrase "any good" to be synonymous with "the top 5% of the applicant pool" (aka the top few people in class).

Perhaps I'm just touchy about this issue because I've seen a lot of journalism in the past few years (which is then repeated by parents, teachers, guidance counselors, etc) that seems to espouse the idea that if you go into software you will get a 6-figure salary by typing on a keyboard all day, and therefore we should push all kids (and adults?) to go into software, because who wouldn't want such a grand salary for working at a computer all day? The messaging surrounding our field just seems highly misleading due to the implication that everyone gets that big salary, which in my opinion is making the situation you describe with "just OK" students worse as more and more duped students muddle their way through CS classes hoping for a pot of gold at the end. I think both the students and the industry would be better if we (meaning society, but also those of us in the industry) set more realistic expectations from the start.

[MilesTeg]

Coding + computer science = 150k starting comp minimum if you're any good.

Only in Silicon Valley where your expenses will be super high as well.

Nah, I helped open an Amazon tech center right there in Minneapolis and they paid just as high.

I don't know why people think it's a joke. Definitely new york, seattle, silicon valley you can get that much. But you don't have to take my word for it. Go on team blind or some other site where people share their comp and you can see people with starting compensation all over the US at over 100k.

I don't think it's a joke, I just don't think it is accurate for most fresh CS graduates.

To be fair to @dividendman, they were referring to coders who are "any good." He didn't say anything about the population of CS graduates as a whole. My experience going through a mid-tier state school showed me that there are a rather large number of people who can acquire a CS degree without developing any particular aptitude for software development. I remember taking one class where we had to do pair projects and there was only one other guy in my class who I thought was "any good" (and he felt the same) so we became partners. That worked out great! (He ended up getting hired by Microsoft, btw). Then for the next round of projects they teamed each pair with another pair to make groups of four, and the pair we got stuck with was literally no help at all. Someone who kind of muddles through their algorithms and operating systems classes will earn a degree, but they might be "just OK" at software development, not "any good." The Big Tech firms try very, very hard not to get stuck with the "just OK" folks. The teams I was on before FIRE always had open positions, and not for lack of applicants.

I think this is actually the crux of my issue with the original statement. If you are interpreting the phrase "any good" as "the top few people in class" then yes, I agree with dividendman that if you are one of those people you can earn 150k+ right out of college easily. However, I don't think that most people find the phrase "any good" to be synonymous with "the top 5% of the applicant pool" (aka the top few people in class).

Perhaps I'm just touchy about this issue because I've seen a lot of journalism in the past few years (which is then repeated by parents, teachers, guidance counselors, etc) that seems to espouse the idea that if you go into software you will get a 6-figure salary by typing on a keyboard all day, and therefore we should push all kids (and adults?) to go into software, because who wouldn't want such a grand salary for working at a computer all day? The messaging surrounding our field just seems highly misleading due to the implication that everyone gets that big salary, which in my opinion is making the situation you describe with "just OK" students worse as more and more duped students muddle their way through CS classes hoping for a pot of gold at the end. I think both the students and the industry would be better if we (meaning society, but also those of us in the industry) set more realistic expectations from the start.

Yep, this. Though I would add that one of the reasons coding is being pushed so much is major corporations have a desire to increase the number of workers so that they can drive compensation down.

Mostly because large corporations are run by morons who don't understand that 1 highly skilled programmer may run 2-3x the average cost, but is more productive than the an equivalently compensated stable of code bootcamp hacks.

[FIPurpose]

Coding + computer science = 150k starting comp minimum if you're any good.

Only in Silicon Valley where your expenses will be super high as well.

Nah, I helped open an Amazon tech center right there in Minneapolis and they paid just as high.

I don't know why people think it's a joke. Definitely new york, seattle, silicon valley you can get that much. But you don't have to take my word for it. Go on team blind or some other site where people share their comp and you can see people with starting compensation all over the US at over 100k.

I don't think it's a joke, I just don't think it is accurate for most fresh CS graduates.

To be fair to @dividendman, they were referring to coders who are "any good." He didn't say anything about the population of CS graduates as a whole. My experience going through a mid-tier state school showed me that there are a rather large number of people who can acquire a CS degree without developing any particular aptitude for software development. I remember taking one class where we had to do pair projects and there was only one other guy in my class who I thought was "any good" (and he felt the same) so we became partners. That worked out great! (He ended up getting hired by Microsoft, btw). Then for the next round of projects they teamed each pair with another pair to make groups of four, and the pair we got stuck with was literally no help at all. Someone who kind of muddles through their algorithms and operating systems classes will earn a degree, but they might be "just OK" at software development, not "any good." The Big Tech firms try very, very hard not to get stuck with the "just OK" folks. The teams I was on before FIRE always had open positions, and not for lack of applicants.

I think this is actually the crux of my issue with the original statement. If you are interpreting the phrase "any good" as "the top few people in class" then yes, I agree with dividendman that if you are one of those people you can earn 150k+ right out of college easily. However, I don't think that most people find the phrase "any good" to be synonymous with "the top 5% of the applicant pool" (aka the top few people in class).

Perhaps I'm just touchy about this issue because I've seen a lot of journalism in the past few years (which is then repeated by parents, teachers, guidance counselors, etc) that seems to espouse the idea that if you go into software you will get a 6-figure salary by typing on a keyboard all day, and therefore we should push all kids (and adults?) to go into software, because who wouldn't want such a grand salary for working at a computer all day? The messaging surrounding our field just seems highly misleading due to the implication that everyone gets that big salary, which in my opinion is making the situation you describe with "just OK" students worse as more and more duped students muddle their way through CS classes hoping for a pot of gold at the end. I think both the students and the industry would be better if we (meaning society, but also those of us in the industry) set more realistic expectations from the start.

I was a top CS graduate at a state school, moved to a MCOL town to work for giant tech corp. Was hired at about 70k. In a COL calculator, this would be about 120k equivalent in NYC or 170k in San Fran. So yeah, I wasn't too hurt to not be making over 100k
Worked there for 3 years and topped out at at about 76/77. I had to make 2 jumps before actually hitting a job that gave me a 20-30% pay bump and then another 15% raise after my first year.

In general I would say anyone that makes 150k to start is:

1. lucky to have been able to specialize in something that really takes off, but thinks it's because they're smart.
2. lives in a HCOL city
3. Is actually a Linus level coder


Now making 150k in Minneapolis is a good deal, and MN is a great place to make a lot of money with low cost (heck, I don't want to live there).

[FIPurpose]

For example I was recently interviewing for a job that was offering 40% higher pay than I currently make, but they very specifically want Kotlin Android experience. Now, I can code Java like any other CS person, and I've tinkered with Kotlin. But tinkering wasn't good enough for this company. They wanted someone with Kotlin-Android experience. So I was supposed to have gotten in on Kotlin being introduced to Android like what 3 years ago? Same thing happened with Ruby on Rails. Companies stuck a lot of their chips on a particular programming framework, refused to hire general CS workers and instead decided to pay Ruby specialists 30-50% more. It doesn't make sense to me. Never has.

Most CS work out there is maintaining and improving tech that is likely already 4-5 years old at minimum. So finding a job that will allow you to develop highly desirable niche CS skills to put you into the top tier of engineer pay is lucky

[PaulMaxime]

For example I was recently interviewing for a job that was offering 40% higher pay than I currently make, but they very specifically want Kotlin Android experience. Now, I can code Java like any other CS person, and I've tinkered with Kotlin. But tinkering wasn't good enough for this company. They wanted someone with Kotlin-Android experience. So I was supposed to have gotten in on Kotlin being introduced to Android like what 3 years ago? Same thing happened with Ruby on Rails. Companies stuck a lot of their chips on a particular programming framework, refused to hire general CS workers and instead decided to pay Ruby specialists 30-50% more. It doesn't make sense to me. Never has.

Most CS work out there is maintaining and improving tech that is likely already 4-5 years old at minimum. So finding a job that will allow you to develop highly desirable niche CS skills to put you into the top tier of engineer pay is lucky

I hate when companies do this. You know Java - you can program in Kotlin in short order. Companies should hire for quality of experience and problem solving not some dumb syntax that might take a couple weeks to get down.

I wanted to learn Kotlin so I decided that our next project would be in Kotlin. Went just fine. I still get people at my company questioning this from time to time since a lot of them are Java developers.

[Alchemisst]

... what you learn will be redundant in a few years (you'll never become an expert over time like other careers such as law, accounting medicine etc)

Why would a coding be any less demanding than law, accounting, or medicine? Every skilled profession benefits from continual education.

A lawyer who doesn't keep up-to-date with recent developments in the legal field is likely to get sued for malpractice.

My mother, (a tax accountant), spends roughly 5% of her time on studying and training/classes to be aware of changes to constantly changing tax laws.

A doctor who graduated 30 years ago and has be coasting ever since would not know about medical lasers, genetic testing, viagra, etc.

That's true, but only to a point I believe. If a qualified cpa/ca accountant or lawyer took say a 5-10 year career break they could potentially walk back in with maybe just having to brush up on the new tax laws etc. However for a programmer/ developer I think they would have a hard time finding a job and potentially have to re learn a lot of totally different technologies

[ctuser1]

That's true, but only to a point I believe. If a qualified cpa/ca accountant or lawyer took say a 5-10 year career break they could potentially walk back in with maybe just having to brush up on the new tax laws etc. However for a programmer/ developer I think they would have a hard time finding a job and potentially have to re learn a lot of totally different technologies

Pace of change is faster. from one (common) perspective.

Those changes, however, would not be so jarring for someone with a good grasp of the underlying science. I have already seen many new-flavor-of-the-month stuff come about - but none of them break new ground in terms of the science behind it. IOC/DI, declarative programming, functional programming - the math of all those were invented 50+ years ago. So, if you have some degree of familiarity with them - you shouldn't have much difficulty picking up the newest fad out there.

Now, CS is a *hard* science if you focus on the science bit. In fact, I'd argue that naming core-"CS" as a separate subject is just a fad. Folks just took some of the more complicated and non-intuitive parts of discrete mathematics, packaged it, and called it CS. Becoming familiar with all the relevant concepts of CS takes much more brain-capacity than becoming an accountant, doctor or a lawyer. Hence, far fewer people will actually have the capability to do that than would be capable of becoming a good lawyer/accountant/doctor etc.

So are you someone who loves those "hard" bits that others find daunting? Then you definitely should learn to code. Heck, you won't have to spend any effort to learn to code.

Are you not (like, maybe 99%+ of the population)? Then - it depends!!

[bacchi]

I was a top CS graduate at a state school, moved to a MCOL town to work for giant tech corp. Was hired at about 70k. In a COL calculator, this would be about 120k equivalent in NYC or 170k in San Fran. So yeah, I wasn't too hurt to not be making over 100k

Same. Much of the COL is about purchasing a house, granted, but I'm guessing living 4 to a room gets old when you're 30.

Pace of change is faster. from one (common) perspective.

Those changes, however, would not be so jarring for someone with a good grasp of the underlying science. I have already seen many new-flavor-of-the-month stuff come about - but none of them break new ground in terms of the science behind it. IOC/DI, declarative programming, functional programming - the math of all those were invented 50+ years ago. So, if you have some degree of familiarity with them - you shouldn't have much difficulty picking up the newest fad out there.

It's not about whether one can pick up the newest fad, though. It's about whether you can convince an employer of that.

In a lot of cases, it doesn't matter if you worked through the tutorial, have used the droid simulator, and can talk about the differences and most important API calls. Manager: "Sorry, it's not Java. You need 3 years professional experience in Kotlin."

We've all been there and maybe it's because there are so many "coders" who only know how to do one thing and where each new NoSQL db takes weeks to learn. Or it's because there are too many b-school hiring managers.

[ctuser1]

It's not about whether one can pick up the newest fad, though. It's about whether you can convince an employer of that.

In a lot of cases, it doesn't matter if you worked through the tutorial, have used the droid simulator, and can talk about the differences and most important API calls. Manager: "Sorry, it's not Java. You need 3 years professional experience in Kotlin."

We've all been there and maybe it's because there are so many "coders" who only know how to do one thing and where each new NoSQL db takes weeks to learn. Or it's because there are too many b-school hiring managers.

I have a slightly different take on this. I don't dispute the facts - just the interpretation or the emphasis.

You don't staff a 10-people team with 9 tech-SME's, or even 5 tech-SME's. At most, a 10 people team will likely have 1 tech SME, or maybe 2 at max. Finding those SME's is often very difficult, especially in new(er) technologies. If you don't find someone who is truly an SME in that very niche technology - then you are better off using one of your existing in-house tech-superstars have him grind through the learning curve. Hence the take-no-hostage approach to the recruitment of these positions. You are looking *only* for uniquely qualified people in these positions, don't care for people who are just "qualified", and you cast a very wide net for these positions because they are so difficult to find.

These positions create a huge amount of splash. But irrespective of the amount of noise they create, they are only a tiny number of total jobs out there. A vast majority of positions are not filled this way.

I have changed jobs externally (i.e. moved employers) 3 times, and moved internally many more times than that (between 2008-2010, when I was hanging on for dear life, I moved internally 6 times in a 2 year period. It seemed like a never ending cycle of find a position, oops thats gone, now find another one!!). My moves have crossed roles left and right (think developer, project manager, management consultant, quant). ALL of them have been done through networking. Not a single one of them was because I was the expert on a specific technology. Heck, in the current job I have, I had never coded in the language I use right now outside of as a hobby to try it out. I was not even in a "technical role" in my old job before I joined this. I was basically a "management consultant". I had a significant amount of domain knowledge, and was hired based on that into a technical role with the expectation that I will take a couple of weeks to pick up this new language while overseeing a small team who all were better coders than I was in this specific language.

Now, my experience is also not an "average case". But the point I make is that vast majority of hiring happen between the two extremes of "tech skills don't matter" to "need uniquely qualified SME". So I am not quite convinced it is as big a problem as it appears in the first glance if you only look at those high-noise jobs on all bulletin boards that are looking for uniquely qualified SME's.

[BDWW]

Just my opinion,
  But I think it's probably common for applicants to lie when applying, and then grind through as much of the learning curve before your start if you get the job. Or be smooth enough to claim most of the qualifications but convince them it's close enough.
  There's no way companies are finding some of the ridiculously specific qualifications I see in some job postings.

edit: My favourite I've seen is "need 5 years experience in XXX" when XXX is relatively new technology and the only way you might have 5 years experience with it is if you were on the original development team.

[maizefolk]

Keep in mind some of those super specific job postings aren’t real, they are a necessary precondition for filing the paperwork for an H1B visa (specifically for bringing in people with specific technical skills that it is not possible to hire domestically).

[Alchemisst]

That's true, but only to a point I believe. If a qualified cpa/ca accountant or lawyer took say a 5-10 year career break they could potentially walk back in with maybe just having to brush up on the new tax laws etc. However for a programmer/ developer I think they would have a hard time finding a job and potentially have to re learn a lot of totally different technologies

Pace of change is faster. from one (common) perspective.

Those changes, however, would not be so jarring for someone with a good grasp of the underlying science. I have already seen many new-flavor-of-the-month stuff come about - but none of them break new ground in terms of the science behind it. IOC/DI, declarative programming, functional programming - the math of all those were invented 50+ years ago. So, if you have some degree of familiarity with them - you shouldn't have much difficulty picking up the newest fad out there.

Now, CS is a *hard* science if you focus on the science bit. In fact, I'd argue that naming core-"CS" as a separate subject is just a fad. Folks just took some of the more complicated and non-intuitive parts of discrete mathematics, packaged it, and called it CS. Becoming familiar with all the relevant concepts of CS takes much more brain-capacity than becoming an accountant, doctor or a lawyer. Hence, far fewer people will actually have the capability to do that than would be capable of becoming a good lawyer/accountant/doctor etc.

So are you someone who loves those "hard" bits that others find daunting? Then you definitely should learn to code. Heck, you won't have to spend any effort to learn to code.

Are you not (like, maybe 99%+ of the population)? Then - it depends!!

CS is harder than becoming a doctor?

[ctuser1]

That's true, but only to a point I believe. If a qualified cpa/ca accountant or lawyer took say a 5-10 year career break they could potentially walk back in with maybe just having to brush up on the new tax laws etc. However for a programmer/ developer I think they would have a hard time finding a job and potentially have to re learn a lot of totally different technologies

Pace of change is faster. from one (common) perspective.

Those changes, however, would not be so jarring for someone with a good grasp of the underlying science. I have already seen many new-flavor-of-the-month stuff come about - but none of them break new ground in terms of the science behind it. IOC/DI, declarative programming, functional programming - the math of all those were invented 50+ years ago. So, if you have some degree of familiarity with them - you shouldn't have much difficulty picking up the newest fad out there.

Now, CS is a *hard* science if you focus on the science bit. In fact, I'd argue that naming core-"CS" as a separate subject is just a fad. Folks just took some of the more complicated and non-intuitive parts of discrete mathematics, packaged it, and called it CS. Becoming familiar with all the relevant concepts of CS takes much more brain-capacity than becoming an accountant, doctor or a lawyer. Hence, far fewer people will actually have the capability to do that than would be capable of becoming a good lawyer/accountant/doctor etc.

So are you someone who loves those "hard" bits that others find daunting? Then you definitely should learn to code. Heck, you won't have to spend any effort to learn to code.

Are you not (like, maybe 99%+ of the population)? Then - it depends!!

CS is harder than becoming a doctor?

The science part can be very complex. e.g. Black Hole information paradox?
https://en.wikipedia.org/wiki/Black_hole_information_paradox#:~:text=The%20black%20hole%20information%20paradox,devolve%20into%20the%20same%20state.

Full disclosure - I don't understand it.

Then there are the more recent advances in quantum computing that I haven't even attempted to understand yet - they sound so daunting!! From all the articles that drop into my email inbox, there seem to be a huge spurt of activity in this field of late - unfortunately well beyond my grasp.

Getting admission to become a doctor is definitely harder. The material of any of the "hard sciences" (i.e. Math/Physics and anything related to them - including CS) are much harder, IMO.

That said, much of the "hard" stuff of CS is not required for programming. You definitely don't need to understand the theoretical implications of Turing Machines to write a computer code, or that of generative grammars or compiler theory to write a regex. Having that understanding would help in trying to make sense of novel situations - but that is hardly a necessity. So - if your focus is just to learn programming and land a job then you don't need to worry about any of these.

I did not realize you are the OP when I geeked out on you on the "rate of change" topic. Please don't worry about the harder CS topics if you are just trying to learn to code. I definitely didn't want to discourage you from that.

[mm1970]

That's true, but only to a point I believe. If a qualified cpa/ca accountant or lawyer took say a 5-10 year career break they could potentially walk back in with maybe just having to brush up on the new tax laws etc. However for a programmer/ developer I think they would have a hard time finding a job and potentially have to re learn a lot of totally different technologies

Pace of change is faster. from one (common) perspective.

Those changes, however, would not be so jarring for someone with a good grasp of the underlying science. I have already seen many new-flavor-of-the-month stuff come about - but none of them break new ground in terms of the science behind it. IOC/DI, declarative programming, functional programming - the math of all those were invented 50+ years ago. So, if you have some degree of familiarity with them - you shouldn't have much difficulty picking up the newest fad out there.

Now, CS is a *hard* science if you focus on the science bit. In fact, I'd argue that naming core-"CS" as a separate subject is just a fad. Folks just took some of the more complicated and non-intuitive parts of discrete mathematics, packaged it, and called it CS. Becoming familiar with all the relevant concepts of CS takes much more brain-capacity than becoming an accountant, doctor or a lawyer. Hence, far fewer people will actually have the capability to do that than would be capable of becoming a good lawyer/accountant/doctor etc.

So are you someone who loves those "hard" bits that others find daunting? Then you definitely should learn to code. Heck, you won't have to spend any effort to learn to code.

Are you not (like, maybe 99%+ of the population)? Then - it depends!!

CS is harder than becoming a doctor?

Depends on what kind of hard.

Full disclosure: I can't program my way out of a paper bag.  I really should take a few classes to help me with my work.  I do a lot of data analysis, and I would be more efficient if I could do a bit of coding.

I'd say concepts in CS and programming can be very hard, where doctoring (depending on the specialty) is not as mentally taxing.  It's very physically taxing to get through med school and internship.  I can understand biology and chemistry and physics, as long as you don't bring in imaginary numbers.