alt Hacker NewsA series of vignettes from my childhood and early career
Comments
I posted this link after reading all the way. Jason actually makes a good point - its just that this title is loud. Blog post itself isn't claiming the death of software engineering at all. If anything, it just shows that every five or ten years someone claims software engineering is dead.
Its not dead at all and it wont die either.
Why? chagpt, or figma or v0 can spin up a few pages of brochure site, even some blog posting level web apps, basic cruds you know. But I don't think it will replace full software engineering.
I work with a large codebase, thats almost 30 years old, multiple framework ( backbone, react, angular) and then java, python for backends. All from different phases and everything is stitched together to make it work, and have a well profit making business going on. There is no model or chatxyz that can dig throug all these connected apps and services and replace our engineering team. It helps us here and there- yeah a lot.
Seems like groundhog day, although I'm not sure I remember anyone telling me that software engineers were on borrowed time until relatively recently and I'd largely ignored it.
Yet one thing does seem different for anyone who just missed the dotcom crash, is that the roles available have fallen off a cliff while the numbers looking for roles seem to be up, at least in the UK. The UAE is even worse. I've spent 20 years hiding from recruiters and now they're all leaving me on read. Karma, maybe.
Isn't this a bit revisionist? I started to become interested in programming around late 90s and I don't remember anyone floating the idea that OOP, libraries or IDEs will make programming obsolete as a profession. If anything, pre-2023 most programmers considered their job as one of the hardest ones to automate.
I was, for a long time, scared of my future due to the low/no-code, automation, LLMs, outsourcing, etc. Until, at some point, I realised something simple - the risk factor for my job is not determined by how good new tools are, but only by how lazy people are about learning and adopting them. And here history gives another lesson - we never learn, eternal cycle of mistakes will continue.
Have we forgotten that COBOL was going to eliminate the need for programmers (even though FORTRAN and ALGOL had both somehow _increased_ the need (not to mention LISP, which for some reason mostly no one did at the time (though it still got more attention than APL)))?
But then BASIC came along and really finished the job. Now _anybody_ could program, without needing all that specialized training. It was just a skill, not a job!
Jason has nailed it. If he were older, his list of vignettes might be longer, but the point would remain the same.
Ahh - so many gems!
"The dream of the widespread, ubiquitous internet came true, and there were very few fatalities. Some businesses died, but it was more glacial than volcanic in time scale. When ubiquitous online services became commonplace it just felt mundane. It didn’t feel forced. It was the opposite of the dot com boom just five years later: the internet is here and we’re here to build a solid business within it in contrast with we should put this solid business on the internet somehow, because it’s coming."
Yes. And it continues on.
Just to react to the "i automated myself out of a Job" part: happened to me at my first job, as we automated more and more our deployment, we could take more and more clients, and I ended up spending 90% of my time fixing routing issues, onboarding clients, integrating their ETLs or inhouse software, or fixing their "chmod -R 777 /" and other mistakes. Which wasn't an issue when it was 30%, or even 50% of my job to be clear, but became extremely boring and soulcrushing at the end.
I'm still happy i automated stuff, that was the interesting part of the job,
Every time a new tool is made, more complicated things come within reach. And then you need a guy who can use the new tool. Who is that guy going to be? He's gonna be the guy who already knows how to use the previous tool.
This should be taken in the context of software demand versus production. Demand has risen steadily over these years and efficiency gains in production just unlocked more demand because it lowered the cost. I think LLM coding will also unlock more demand but only up to a point until unit cost is so low it doesn't make sense to have a human in the loop anymore. Then we software devs are SOL.
There is one and only one important question... have companies been hiring as many juniors as in the past recently?
> New problem sets not covered by the garden path come up all the time.
I believe the appropriate term is “happy path”, not “garden path”.
> “Coding is over, with Object Oriented programming one person who is much smarter than any of us could hope to be will develop the library just once and we will all use it going forward, forever. Once a problem is solved it never needs solving again.
This kind of happened, and it was a good thing.
Have to say, love the typography on this site. Did some quick digging and the headings font is an Adobe font called Forevs, with a similar Google font being Fraunces.
Not sure what to make of the post.
Software Engineering isn’t a profession. Software Development is. Software development as a profession may wane and morph, due to advancements in technology and other creations.
I don’t know any engineer who has ever said “engineering is a dead end”. Because that’s an obviously nonsensical statement. So, engineering stands on its own for time immemorial.
And no - I’m not nitpicking over terminology. Learn engineering.
I still love programming. Even more so after trying out llm coding in some projects.
I thought this was Jason Schreier and got confused for a minute.
I’ll be honest, I’m confused by the ending. It says this is a series of jabs at LLM hype, and I understand that it’s intended as a series of jabs against those who would say “software engineers will be out of a job”, but I’ll be honest I think only a loud minority are saying that yet it’s treated like it’s a majority position.
All the stories listed seem interesting, but none of them seem all that relevant.
I feel like most people understand that this is a seismic shift in abstraction layer, but intelligent people will still be in demand to manage the machines at whatever level is currently highest. The motor car didn’t kill taxi drivers, unless those who drove a carriage refused to learn how to drive a motor car.
Perhaps I’m not expressing my point very well… but this feels like both an argument against something almost no one is saying seriously, and it uses examples that also aren’t that applicable to the current situation other than having the commonality that people have said before that software engineers will die out. Make me wonder… How many times did people think an invention would kill off a job incorrectly, until one day it actually did?
Intelligent and well educated people will always be in demand somewhere. Until we’re in some post money utopia, we’ll just have to roll with the punches. In the meantime, HN readers like ourselves will simultaneously upvote any article that says humans are super necessary down at lower levels of abstraction and are way better at coding than LLMs, whilst quietly also coding less and less by hand and crawling up that abstraction layer themselves. That’s just human nature.
The ongoing issue is the maintenance.
This can't be solved without fully trusting the LLM period.
Just don't autopilot on important code you want to own. That's good start.
The title is “The Death of Software Engineering as a Profession", why was this changed? Not really fair to the author.
The reports about our death were greatly exaggerated.
I really like the font this author is using for the headers. Reminds me of the fonts used on covers of 1970s scifi novels.
[dead]
[dead]
Forgive me but a lot of the examples seem like strawman.
> The dream of “multimedia” became commonplace and everyone just accepted it as normal. I’m not aware of any industries that collapsed dramatically due to multimedia.
But "multimedia" was never purported to be something that would lead to collapse of any segment of the industry, much less industries. If anything, the multimedia hype was purported to increase IT work which it did for some years.
> In 2000 a coworker took me aside and showed me his brand-new copy of IntelliJ IDE. “It’s over for us,” he said, “this thing makes it so programmers aren’t strictly necessary, like one person can operate this tool and they can lay the rest of us off.”
I've a hard time believing this. Literally nobody I've met was ever mistaken that IntelliJ would mean the doom of software engineering work. It's a great IDE and all IDE including IntelliJ required engineers to write code with them. Nobody was foolish enough to really think one engineer or one manager or one salesperson can "operate" IntelliJ and generate all the code to meet business requirements.
> And then he showed me the killer feature “that’s going to get us all out of a job:” the refactoring tools.
I'll bet there was no such "coworker". No sane person would think "refactoring" could mean "magically understand business requirements and write code"? All of this sounds like strawman setup so that the author could go on to making their next point like the bit where he challenged his "coworker" and asked if refactoring tools can write new code.
Don't get me wrong. The rest of the post is on money though. I just think the post would do better without these fake stories to set up strawmans only to take them down. Feels a bit forced!
"The Death of Software Engineering as a Profession" ... is greatly exaggerated.
Could be the title of the piece.
I agree: throughout my own career as a programmer (I prefer the more blue-collar sounding term—it better fits my skill set) I have also seen large changes in the industry that certainly made waves, did not capsize the profession.
At the same time, the profession I retired from was by no means the profession I entered into in the '90s. I confess I liked the older profession better.