Professor gave us dire warning that CS is going to be the next Art History major | Hacker News Reader

Professor gave us dire warning that CS is going to be the next Art History major

reddit.com

51 pointssizzle49 comments

There's some good points here. I don't have a degree and, as a senior in their late 20s, find myself teaching things I consider basic to CS graduates.

I work in one of the big 5. I keep finding myself wondering why we have 50+ engineers on our team when we can never deliver anything on time. Everything we build feels cumbersome, over-engineered, and overly-reliant on bespoke internal tooling that nobody outside of our company uses. After coming from startup land and building infinitely more useful products in a 10th of the time with 1-pizza teams, I wonder how the hell this gargantuan company can move around $1B market cap on wall street when its so inefficient and full of incompetent engineers.

It genuinely feels like we've reached peak too-many-cooks. I keep expecting mass layoffs to cull the fat but they never come. We just keep hiring more and more and more.

This can't be sustainable.

If you don’t mind saying, what company do you specifically work at currently?

Sounds similar to my situation. I try and put it down to programmers like programming to justify why the devs here take the most complex route rather then the cheapest/fastest/reliable route.

I've worked at several of the big five, as well as a couple smaller companies, and I can't say I've had a similar experience. As long as people are willing to deprecate, delete, and consolidate continuously like your sanity depends on it, the scope of problems seems to be constantly shifting as things improve.

There will always be problems. Every time we solve one thing, we move to a place where we discover new kinds of issues. The question is whether we keep running into precisely the same kinds of issues over and over again, and in my experience, this has not been the case.

I’ve heard this type of speech before. It’s a very candid almost scary talk designed to shake you up and get you focused. The gist was don’t think this degree is all you need. If you get lax some jerk with a history degree is going to take your job and you’re going to flying airplanes full of rubber dog poop from Hong Kong (sorry for adding the top gun reference... for me everything boils down to 80s movie quotes). And these speeches are right - you can’t rest.

Every couple years someone would predict software developers (where CS majors were heading) would be out of job for various reasons. Spreadsheets would let business people to do analysis. Report writers let people to generate reports. Visual programming let people drag and drop to build software. VBA will take care of all business needs. Rule engines will generate programs en-mass. Code gen from UML/State-Diagram would make developers obsolete. The latest is the claim that AI will generate software and replace developers. These people are clueless about the nature of programming, software abstraction, and the trends to automate all aspects of life. Software and developers are more likely to put people from other fields out of job than the other way around.

It's more likely the teaching aspect of the college professor will be replaced by AI/machine in the near future.

lol. Yeah, if schools can commit to it, top notch online lectures from top notch schools, will surely turn not-so-good-profs into a tenured TA.

Yeah, I remember the big art history boom times. It was crazy, it seemed that everyone was getting into the space, salaries seemed to be going up almost daily, the completion was cut throat.

The professor should stick to teaching computer science and not try to be market economist.

I thought we already went through this in the early 2000's when everyone predicted all CS work would be outsourced to India in the near future. The fact is we work in an industry with potentially crazy profit margins, a handful of skilled programmers can create something worth X1000 their income. Skimping on your development team has been proved again and again to be a terrible idea.

I do agree that if you are in it for the money, you will have a bad time and not get very far.

>> a handful of skilled programmers can create something worth X1000 their income.

What you're trying to say here is that 'only a handful of programmers have been at the right place at the right time to earn the position that gives them enough equity worth x1000 the income of others'.

There are many talented developers out there.

Trying to say that an average developer from India has the same chances of getting involved with a highly-financed startup at it's early stages as an average developer from Silicon Valley is just misguided at least.

India aside, we have the same issue with zip codes in America. The number of high school students from Santa Clara County to become developers or work in lucrative tech jobs dwarfs the number of students from similarly populous counties from other states landing the same jobs. Colocation, mentoring, networking, and cultural similarities really matter.

The real problem with CS is that it changes quickly and skills become obsolete. If you don't actively work to maintain current skills you are likely to end up getting turfed out in the next downturn. Even that's not a unique problem. IBM assembler programmers and blacksmiths are both pretty rare these days, for similar reasons.

Computer Science - as in the fundamentals and mathematics behind computation do not change frequently at all. Graph theory never really goes out of style, and the biggest changes have been with respect to distributed computing and conflict resolution.

The thing that scares me the most is the confusion of software frameworks and platforms with "CS" and fundamentals. Its as if people believe there aren't any fundamentals at all.

I mentor students in CS programs. A lot of the material is very topical. Also, the stuff that gets you a job (or helps you keep it) is likewise topical. It's fine to know about dining philosophers but if you can't implement solutions in the language du jour a lot of places won't hire you.

I really think that explains much of the problems of our industry. CS should teach the fundamental concepts all these languages are based upon. When you've learned that a new language is merely a different mapping of these concepts rather than something completely new, it becomes quite easy to pick up new things. You understand they are just assigning different syntactical weight to the same old concepts.

In that respect I liked that my education taught me things like LISP and smalltalk that are quite different from C/C++/Java that was big at the time. If I didn't have that field of view I might think smalltalk is its own thing entirely and I'd have to relearn everything.

This is a point where reasonable humans can disagree. There's a middle ground that includes enough current practice to get you an entry level job + enough of the theory to ensure you can continue learning efficiently over the course of your career.

But IBM Mainframe programmers aren't as rare as one might think. And that platform is one a young person can grow with their whole career instead be whip sawed by the latest Javascript framework every year.

Sure, ASM is a thing for some people. But such programmers are a vastly smaller fraction of the programming pool than they were in the days when application packages like accounting systems were written in Assembler. COBOL, FORTRAN, and PL/1 put an end to that.

In the future there will be three jobs: artists, engineers, and sellers. Artists will design and create the products but they’re at the top of the pyramid. Next you have sales for people who don’t have the technical or creative skills but who know how to hustle. Then you have the people who are going to implement the ideas of the artist and give sales something to sell. Insofar as CS is how we build with information I see this as the biggest class and see no oversupply, only oversupply of unqualified talent.

What about just ordinary proles? I mean the big shiny economic machine is not nearly so automated as it appears to be. It’s a mechanical Turk composed of billions of low paid laborers. Even Elon musk is discovering that there’s a limit to automation.

If you are doing creative handmade software (bespoke) then you are doing an artistic endeavor, like writing, solving math problems via proofs. I've been fortunate to do that during my career in cs - working on infrastructure software. But how many databases does the world need? Eventually we'll have enough of these tools and more people imho will just put tools together to solve problems.

I agree 100%. CS programs are packed now because they pay is so high vs other majors. The problem is that there are only so many positions the economy can accommodate and many of the students going through the system now will never get a job in the field.

The outstanding graduates from the prestigious programs will always get a great job but everyone else will have to fight for the jobs that are left.

Look at India. The schools are turning out a bunch of CS/Engineers but the economy can not accommodate the number that are graduating so many of them are getting jobs they could have gotten without all the education. The same thing is happening here.

The real winners will be the ones that use CS as a stepping stone to creating a business, therefore, creating your own need for the degree.

If I had to get a degree now I would pick a combination CS and business degree with a minor in a field that's in the humanities and start to figure out how I'm going to put it to use.

A CS degree alone will lead to a job that does not need it and could have been gotten with any college degree. It's really supply and demand in effect.

We definitely do not meet the demand for programmers in the market. Big software companies don't only look at the number of CS grads since the real need is for talent. Not all of these graduates have the skill required to work in top tier companies, but I'd argue that not all jobs need top tier talent, and that there's plenty of jobs out there that are still looking for the average joe programmer.

Look at this page: https://www.bls.gov/ooh/computer-and-information-technology/...

Programming jobs are expected to fall. But I bet people seeking and training for jobs will continue to increase.

Also, you emphasize my point. Joe programmers don't need a full CS degree. From experience, I know a 2-year associate degree is enough.

There are so many dev jobs in the big coastal cities. Seattle must have 20k openings between just the fangs, 1000+ little companies that each want 10 more devs (including mine). And in big cities like Dallas or Houston there is just a lot of demand. 35k grads in 2016 in the entire us, and Seattle could probably about absorb them all if you add in the east side. there's never been a boom like today for lawyers - its just not comparable.

1. without more context about who the professor is targeting his tirade at and which company C levels he's actually talking, it's hard to evaluate this

2. will there be shrinking in the middle band of SAP/Infosys companies where work has already been proven to be fungible (by their h1b shenanigans) in the near future? probably. does this sort of pressure exist in the middle band of any profession? yes (even if the AMA is fighting tooth and nail to artificially limit supply in the case with MDs)

3. is AI mostly a hype-add to boost valuation? yes. but does a lot of it work and generate real revenue? yes. almost by definition the more valued something is, the more likely it is overvalued. I fear this professor hasn't seen most of the unsexy AI work. where 90% of the value may come from the actual AI work, but 90% of the work is on the infra around it.

I'd be more likely to pay attention to someone who'd already laid down, say, a $100 million bet on this idea.

AI's are still running cars into freeway barriers and firetrucks. The 'big 5' can't even hold on to "their" own data (yours).

> He made the argument that globalization (read: Asia) is pumping out masses of CS people who corporations will continue to use to write cheap code and do CS work at low ass wages.

Haven't we been through that before?

This pycon talk from 5 years ago, is still, extremely relevant

https://youtu.be/_xpKFWzMceE

More computer programmers of low quality means there will be even more work for programmers. Oil engineers don't create work for themselves.

Yes there are not that many good people in CS. Nothing new here.

But still we can't find enough good people in munich at all. And it is not just munich.

But the reason might be ... Munich? I certainly wouldn't want to move there.

What is bad about Munich? (For those of us off the continent)

Good question. High salaries, clean air, very save (one of the savest cities in europe), etc.

Not that much going on party wise and not that easy to find flats.

Cost of living too high.

I've currently plenty of space and a half hour commute. That wouldn't be possible in Munich even if I got paid much more than now.

It's the same as Berlin. So...

Is that you, Tableau Munich office? :-)

Seattle can't find enough programmers either.

Anyone can major in Art History without a lot of effort. It’s not a differentiating credential for many jobs.

CS is a very difficult major. It’s a differentiating credential for many jobs. That’s why the market clears differently.

The quality of CS programs (and grads) may go down but there will always be a market for the talented. I think Finance and Accounting might be the more appropriate comparison. There is a lot of outsourcing and crap accountants, but the best financiers still do very well.

would sign that. Basically what you learn in a good CS program isn't trivially accessible or more or less "soft" knowledge requiring only a (very good) command of the language you already need for day-2-day communication. It requires logic and strict thoughts – while those things aren't necessarily making you a good coder, this are skills not wielded by the masses anytime soon because they require (for most people) to really sit down and learn.

For art history you probably go there, look into your textbook and just start talking about random stuff associated with limited data and if you're audience isn't really into their code/peer-bias-bubble/.. there is a big range of "reasonable" things you can say. (saying that as a physics major with a lot of CS friends and random acquaintances which I regularly impress by just introducing random facts (they don't think a STEM-guy knows...) in conversation and giving them a techy twist...)

As to the salary question: apparently a lot of people see academia as a sort of fancy tradeschool with guaranteed above average income and of course this is not the main objective of academic ventures and if everyone thinks that, some has to be the below-average-guy or gal. For the average person studying solely for the purpose of eventually making more money is just plainly stupid.

>It requires logic and strict thoughts

>art history you probably go there, look into your textbook and just start talking about random stuff associated

(emphasis mine)

>with limited data

This is all really funny stuff.

I don't really get your comment. Yes, I didn't write any assignments in art history nor do I exactly know, what you actually learn from art history, but guessing from checking other peoples assignments, talking to people and looking at their lecture notes in similar field (e.g. communication science, ancient languages and archaeology) I assume that you basically have to read up some facts and then can go to any lecture you fancy and start contributing in a more or less meaningful way. You can't do that in CS or physics or engineering – it just won't work.

Note that I don't intend to belittle those kind of studies; if you put a lot of effort into it, you're most certainly reaching levels of conceptual understanding and a command of language I might never achieve in my life. If one gets "satisfaction"/"gratification" [whatever you need to keep going on a personal level (excluding modern day BS where these things are just used for justifying the continuing exploitation of labor)] from that, I'm all up to you pursuing it. Still, looking at the kind of person who idles similar studies while he/she's waiting tables at Burger-King, the whole idea of academic studies seems a little bit pointless (similar to the CS guys rushing college for the best paying IT-project management job) – though you probably also get to learn a thing or two about the way we live (contrary to the drones going for the highest-paying majors)

My comment was basically pointing out the irony of claiming that art history people say stuff with limited data, based on your obviously limited experience of art history as a field.

> I assume that you basically have to read up some facts and then can go to any lecture you fancy and start contributing in a more or less meaningful way.

This isn't true, unless you're an undergrad at a really bad university. The really 'unproductive' fields like art history tend to have a pretty asymptotic difficulty curve, since funding is almost impossible to find. That makes later stages really, incredibly competitive, unless you're basically willing to pay through the entire process.

One thing to note about hierarchy of perceived difficulty of university courses, is that it almost always maps to how useful these courses are to industry, and almost never maps to how complex the subject actually is. Traditional Chinese literature, for instance, is horribly difficult. Computer science, on the other hand, is a subject where people actually try hard to be understood, and don't write using in bird-worm seal script. Nonetheless, if you are a computer scientist, people will assume you're smart, and if you're studying chinese literature, people will assume you're a goof.

>Anyone can major in Art History without a lot of effort.

Another day, another Hacker News comment asserting the inferiority of any non-technical subject based on zero experience or evidence.

I don't think we need evidence, it's like common knowledge. Medicine, engineering or CS students don't have really time to party around like the art history, journalism and liberal studies folks do.

That is the cliche that I learned in college too, but it seems to be true. Could it be that history and business majors are the smart one because they still do lots of hard work (like 'us' math people) but do it fast enough to party a lot? Yeah, no.

Feels about right to me. But it’s easy to be wrong when it comes to predicting the future. Still, feels about right.

Can somebody explain what exactly happened to petroleum engineering?

Boom and bust. Oil price goes up. The need for gas exploration and drilling goes up. Money pours in. Petroleum engineer's pay goes up. Oil price goes down. The trend reverses.

I still see pretty high pay (~$250k) for petroleum engineers.

Reply on news.ycombinator.com