Ask HN: The collapse of web programming jobs
21 comments
It’s a game developer saying his kind of skills will be in demand while “those other” skills won’t be. I have no idea what will happen, but color me skeptical on this take. If for no other reason there’s not a lot of detail on why.
Specifically why does web programming get easier but game programming does not? Why wouldn’t all programming get gradually easier.
In any case I do think you do yourself a disservice if you can’t pivot to a different part of software. When I graduated college 20 years ago the trend was monolithic Java or C++ apps. There was no Stackoverflow. There was barely a functioning web search engine. AI was rule based. Deep learning wasn’t readily used, nor was machine learning. JavaScript was barely used.
If you’re not constantly thinking about your skills investment and direction, you could be left behind.
Specifically why does web programming get easier but game programming does not? Why wouldn’t all programming get gradually easier.
In any case I do think you do yourself a disservice if you can’t pivot to a different part of software. When I graduated college 20 years ago the trend was monolithic Java or C++ apps. There was no Stackoverflow. There was barely a functioning web search engine. AI was rule based. Deep learning wasn’t readily used, nor was machine learning. JavaScript was barely used.
If you’re not constantly thinking about your skills investment and direction, you could be left behind.
His point isn't that web development is or will be so easy that it will be easy to hire for and thus not in demand. His point is that, most of people who work on "web" code do pointless busywork or just don't work very hard. He believes that this situation will fix itself up eventually, when economics of this amount of waste will become untenable (e.g. when big tech will stop making easy money and will have to actually look at costs).
Him being a game developer, saying that game development will survive, while web programming will die, definitely does project some bias. But I would like to make the argument that web-development as we know it will collapse in the future. Not too far from here, it won't be necessary to have specialized knowledge of languages and frameworks to create applications. You just need to know how to throw configs together. This predates even AI, where its already possible with no-code and low-code platforms to build stuff.
I'm sure that there will be a place for web-development as we know it today, but it probably will not as popular, neither as much of a lever when it comes to building applications and services on the internet. I feel like we were on this path already before AI (no-code, low-code, etc). But with AI it feels like we are on an even faster projection of abstracting most code away behind building blocks. Give it a few years.
I'm sure that there will be a place for web-development as we know it today, but it probably will not as popular, neither as much of a lever when it comes to building applications and services on the internet. I feel like we were on this path already before AI (no-code, low-code, etc). But with AI it feels like we are on an even faster projection of abstracting most code away behind building blocks. Give it a few years.
As a 15 year full time JavaScript developer I think Blow is 100% correct.
Seriously consider what would happen to these jobs if tomorrow things like Vue, React, Angular, Express, Spring (Java), and so on all went away. Most of these people become instantly irrelevant even though demand for products/development in this space will remain constant. Those people are tool users, like screwdriver users who call themselves carpenters. WTF, why and how did we get here?
Programming is programming regardless of whether you are making AAA games, building a web application, or guiding rockets to kill people. Programming is a skill. Employment, however, is not so straight forward. Employment is a game in the same way "musical chairs" is a game and people will chase money irrespective of skills. Games are meant to be won.
The problem is a supply/demand problem. Historically the demand for "web work" (however you want to define that) has been high but the necessary preparation has been astonishingly bad. Most developers are trained by universities to write in something like C++, Java, or Python. Employers solve for this not with training (because that costs money and developers cost money) but with abstractions. The result is a platform that scares the shit out of the people hired to write for it. This is the inefficient that Blow is describing and its very real.
As an example consider the DOM. The DOM is the compile target of the browser. If you are working on the front end that is your foundation and everything layers off of it. Yet, the DOM is a tree model and tree models irrationally scare the shit out of people. The level of hysteria this generates defies the imagination. Most developers would rather be in zombie apocalypse combat than writing to the DOM directly. It's weird.
Another example are JavaScript classes. Why would you ever need classes if JavaScript has native lexical scope, functions as first class citizens, and a scalable object system? Maintainers of the language thought the same thing and were extremely resistant to add classes to the language, but classes were demanded in the ES6 specification almost more than all other features combined. The demand exists because people were/are educated to write code in one way and could not imagine writing code without that one way of doing it.
Seriously consider what would happen to these jobs if tomorrow things like Vue, React, Angular, Express, Spring (Java), and so on all went away. Most of these people become instantly irrelevant even though demand for products/development in this space will remain constant. Those people are tool users, like screwdriver users who call themselves carpenters. WTF, why and how did we get here?
Programming is programming regardless of whether you are making AAA games, building a web application, or guiding rockets to kill people. Programming is a skill. Employment, however, is not so straight forward. Employment is a game in the same way "musical chairs" is a game and people will chase money irrespective of skills. Games are meant to be won.
The problem is a supply/demand problem. Historically the demand for "web work" (however you want to define that) has been high but the necessary preparation has been astonishingly bad. Most developers are trained by universities to write in something like C++, Java, or Python. Employers solve for this not with training (because that costs money and developers cost money) but with abstractions. The result is a platform that scares the shit out of the people hired to write for it. This is the inefficient that Blow is describing and its very real.
As an example consider the DOM. The DOM is the compile target of the browser. If you are working on the front end that is your foundation and everything layers off of it. Yet, the DOM is a tree model and tree models irrationally scare the shit out of people. The level of hysteria this generates defies the imagination. Most developers would rather be in zombie apocalypse combat than writing to the DOM directly. It's weird.
Another example are JavaScript classes. Why would you ever need classes if JavaScript has native lexical scope, functions as first class citizens, and a scalable object system? Maintainers of the language thought the same thing and were extremely resistant to add classes to the language, but classes were demanded in the ES6 specification almost more than all other features combined. The demand exists because people were/are educated to write code in one way and could not imagine writing code without that one way of doing it.
Neh... web programming is not going anywhere. Sure, the current environment is exposing all of the companies that never really had a business model other than raising more and more money. But there's still plenty of space for software to grow within existing profitable companies and new companies that can be more capital efficient.
He claims you can strip out some of the inefficiency which is related to interoperating software... that's not going anywhere. There will alwyas be walled gardens or code that breaks. Maybe that'll even increase. And even though what AI can already do is incredible, I think it'll take a long time before it can build and maintain large scale projects. In the meantime, it'll certainly make developers more efficient, but so have countless other tools.
That said, I think that remote workers are going to have a reckoning. If i'm hiring for people to be in the office, then i have a much smaller pool, replacing people is tougher, etc. But once I'm opening the doors to remote work, why not do so in a country with a much lower cost of living? I see people on HN constantly talking about how to succeed remotely, you need to be able to work in an async environment, etc. Well guess who loves the idea of an async environment? Upper managers who see ballooning payrolls and realize that an async/remote workplace doesn't need to consist of people in high cost countries.
He claims you can strip out some of the inefficiency which is related to interoperating software... that's not going anywhere. There will alwyas be walled gardens or code that breaks. Maybe that'll even increase. And even though what AI can already do is incredible, I think it'll take a long time before it can build and maintain large scale projects. In the meantime, it'll certainly make developers more efficient, but so have countless other tools.
That said, I think that remote workers are going to have a reckoning. If i'm hiring for people to be in the office, then i have a much smaller pool, replacing people is tougher, etc. But once I'm opening the doors to remote work, why not do so in a country with a much lower cost of living? I see people on HN constantly talking about how to succeed remotely, you need to be able to work in an async environment, etc. Well guess who loves the idea of an async environment? Upper managers who see ballooning payrolls and realize that an async/remote workplace doesn't need to consist of people in high cost countries.
The comment about remote workers is nothing new; think of the effects this has had in the past 20 years alone.
Businesses are just now starting to understand that hiring people from overseas is not necessarily the best move. Better to hire in your time zone where it’s easy to communicate, than it is to rely on a constant stream of email messages.
And yet, this still tends to be a problem. One of my jobs has that issue of communication, and it’s basically made me shift from programming to managing the boss and the project.
So… no easy answers to any of this.
Businesses are just now starting to understand that hiring people from overseas is not necessarily the best move. Better to hire in your time zone where it’s easy to communicate, than it is to rely on a constant stream of email messages.
And yet, this still tends to be a problem. One of my jobs has that issue of communication, and it’s basically made me shift from programming to managing the boss and the project.
So… no easy answers to any of this.
If you consider web based SaaS companies then I don't think AI is a reason for collapse, for two major reasons:
1. AI popularity surge is happening at the same time as interest rates going up by a lot. Interest rates impacts borrowing and funding for companies, in many ways money is tighter and cash-flow is more important so borrowed funds with multi-year ROI is not a good idea. These (AI and interest rates) are completely unrelated so correlation (layoffs and software jobs contraction) doesn't mean causation (AI being a cause == not likely, interest rates == maybe, too much hiring during pandemic == more likely). A good question to ask is how many companies have replaced software devs with AI.
2. If you're a SaaS company your core product is software, are you going to let an AI service generate your core product or even submit pieces of your source code to it and feel 100% safe that your code won't be leaked out to other AI users? We can't get humans to document requirements for a feature in a way that can be reliably used to create software, I doubt using AI to generate software is going to work out well. With a SaaS you're betting your entire business on it.
3. Minor nitpick here - the comment is by a game developer. There's a yt video out there of a game developer having ChatGPT write a game. If anything I would think a collapse would happen more so in entertainment industries (as is already with Screen Writers Guild right?) because those are shorter project time-frames with tighter feedback loops to see if its profitable to use AI (and hollywood/game devs re-hash old ideas a lot anyhow but I digress.)
1. AI popularity surge is happening at the same time as interest rates going up by a lot. Interest rates impacts borrowing and funding for companies, in many ways money is tighter and cash-flow is more important so borrowed funds with multi-year ROI is not a good idea. These (AI and interest rates) are completely unrelated so correlation (layoffs and software jobs contraction) doesn't mean causation (AI being a cause == not likely, interest rates == maybe, too much hiring during pandemic == more likely). A good question to ask is how many companies have replaced software devs with AI.
2. If you're a SaaS company your core product is software, are you going to let an AI service generate your core product or even submit pieces of your source code to it and feel 100% safe that your code won't be leaked out to other AI users? We can't get humans to document requirements for a feature in a way that can be reliably used to create software, I doubt using AI to generate software is going to work out well. With a SaaS you're betting your entire business on it.
3. Minor nitpick here - the comment is by a game developer. There's a yt video out there of a game developer having ChatGPT write a game. If anything I would think a collapse would happen more so in entertainment industries (as is already with Screen Writers Guild right?) because those are shorter project time-frames with tighter feedback loops to see if its profitable to use AI (and hollywood/game devs re-hash old ideas a lot anyhow but I digress.)
Can someone more knowledgeable explain the economics of how all this will work?
If the current outlook is that AI will replace programmers, then many other office jobs will be the first or not far off. At that point, who are these corps selling to if there middle and lower class is all but wiped out? Without a steady velocity of money, where is the revenue going to come from that justifies production of goods and services even if fully automated?
Even with an implementation of UBI, all companies will be fighting for scraps because everyone is broke with only enough money for subsistence. Why would anyone have any money in the stock market at that point?
If the current outlook is that AI will replace programmers, then many other office jobs will be the first or not far off. At that point, who are these corps selling to if there middle and lower class is all but wiped out? Without a steady velocity of money, where is the revenue going to come from that justifies production of goods and services even if fully automated?
Even with an implementation of UBI, all companies will be fighting for scraps because everyone is broke with only enough money for subsistence. Why would anyone have any money in the stock market at that point?
I had this exact thought. https://news.ycombinator.com/item?id=35575923
That’s a complex question. You are asking about general equilibrium effects: where the economy will be once companies and workers across industries all adjust to large-scale automation.
Folks who work in roles that are automated away could find employment in other, presumably less well-paid, roles. Or they could end up being the technicians and engineers who manage the automated processes.
The companies can remain profitable even if the purchasing power of workers declines. Their cost could fall due to automation, and their price mark-ups could increase if they manage to drive competitors out of business. So those who own the companies that survive could do well in all this.
Folks who work in roles that are automated away could find employment in other, presumably less well-paid, roles. Or they could end up being the technicians and engineers who manage the automated processes.
The companies can remain profitable even if the purchasing power of workers declines. Their cost could fall due to automation, and their price mark-ups could increase if they manage to drive competitors out of business. So those who own the companies that survive could do well in all this.
In my opinion we are facing the crash of programming gold rush, especially in web development. The only exception is AI (especially LLMs). For a long time, software development was greatly overhyped as a way to make money, with no shortage of self-proclaimed tech leads selling training on how to pass the FAANG interview and common whinings about supposed developer shortage. What goes up must come down. Right now, the market is correcting itself, as the rivers of VC funding for the WeWorks of the the world are running dry under the changing macro-economic conditions.
Web will still exist for a long time and people will keep making money in software development, but things are changing. Easy money is going away, which makes us face the ever-present matter of actually creating the value for the world that was so easy to forget when money was pretty much growing on the branches of Merkle trees.
Web will still exist for a long time and people will keep making money in software development, but things are changing. Easy money is going away, which makes us face the ever-present matter of actually creating the value for the world that was so easy to forget when money was pretty much growing on the branches of Merkle trees.
I agree with Jon Blow, and Techlead has also interesting take on software development as a whole...
multiple forces at effect. investment drying up, the strong assumption AI will replace coders in couple of years, nobody giving a damn about apps because everything interesting is on tiktok / instagram and other social platforms.
the current big thing is content and clout and even that will be replaced by AI in no time. nobody knows what comes after that, star trek or terminator, take your guess...
multiple forces at effect. investment drying up, the strong assumption AI will replace coders in couple of years, nobody giving a damn about apps because everything interesting is on tiktok / instagram and other social platforms.
the current big thing is content and clout and even that will be replaced by AI in no time. nobody knows what comes after that, star trek or terminator, take your guess...
Aren’t game programmers just using something like unity and writing scripts? I feel like that’s probably getting automated away. At the end of the day video games are a corporate beast that prints money. Indie games are like 1% and most people including me have never played them and don’t want to.
Very few are actually making unity or “game engines” from the ground up, like Jon Blow.
Very few are actually making unity or “game engines” from the ground up, like Jon Blow.
> Aren’t game programmers just using something like unity and writing scripts?
Hahaha wow. It’s like saying:
“Aren’t web programmers just using something like chrome and writing scripts?”
Hahaha wow. It’s like saying:
“Aren’t web programmers just using something like chrome and writing scripts?”
the whole thing was meant to be sarcastic lol sorry
I mean, kinda?
Er... yes? :)
Mr. Blow is right about some things and wrong about other things. Just like anybody else.
The fact that he has created a few successful games doesn’t make him an authority on the future of the job market for web developers.
The fact that he has created a few successful games doesn’t make him an authority on the future of the job market for web developers.
Any web programmer can be an AI programmer, because most AI programming is web programming + prompt engineering, and 99.9999% of prompt "engineering" is an absolute farce.
There’s still a bottomless ocean of B2B companies who need web products and services building, that’s not going to go away any time soon. For every planet scale cloud system there are thousands of on premises or basic cloud based applications being developed by people using traditional backend systems.
The job market for a Django + react fullstack dev is not really going to change overnight.
The job market for a Django + react fullstack dev is not really going to change overnight.
A great deal of the software industry as we know it, definitely including webapps and websites, is a ZIRP. Now that money flow is drying up, that industry is withering (unfortunately for me, you, and many others).
[*] https://youtu.be/yodWEPgn8NA