Ask HN: Are web developers on the chopping block?

Here's how you can convince yourself that web developers are still valuable in the face of automation, all the while learning a new skill:

1. Get yourself set up with a machine learning framework e.g Torch or something like that

2. Read a few articles to get familiar with how recurrent neural networks work - I recommend http://karpathy.github.io/2015/05/21/rnn-effectiveness/

3. Train your RNN on Ruby / Python / Javascript code samples.

4. Look at the garbage output despite the working input.

Web developers will be needed for a long time. I'd say we are looking at 20+ years before any automatic development algorithm could replace the coding chops, not to even mention the architectural and operational (security, hosting, etc) skills, needed for modern web development. People rag on webdev all the time as though it's somehow below AI / AR / VR (personally I think AR/VR is the next hype bubble, as no one has found a decent use for it besides games), but even if its just CRUD it takes skill, thought, and competence to make it worthwhile for you and your clients.

Additionally, you could jump ahead of the game by creating a utility or framework that makes web development even easier and then monetize that.

Yes and no.

Yes: A lot of web-development has already been automated for simple sites. Think of how many more web-development jobs would need to exist if Wordpress and Squarespace and other CMS systems weren't powering so many blogs and simple crud apps.

No: Unlike other workers who face being automated out of a job, web-developers are the people doing the automating. Wordpress and Squarespace are being built by developers, after all. In order for my job to no-longer exist, some other engineer has to come in and build something that replaces me. So if you want to be future-proof, all you need to do is make sure you're the developer doing the automating, instead of the developer being automated.

In short: it's dog eat dog. Keep eating other dogs and you'll be fine.

Spot on.

The risk to today's web developers is something that - most realistically - makes web developers so much more efficient that a few people can un-employ many of their fellows. Wordpress and ThemeForest (and their predecessors) did this to the Perl::CGI programmers and HTML jockeys of an earlier age, and e.g. React/Bootstrap/Parse are trying to do this to today's jQuery/Rails experts.

For now, there's plenty of demand if you (can) upskill - those "DHTML experts" who jumped to Rails are still fully employed creating more complex sites (rather than more sites.)

As an alternative to just trying to be better than your peers, patio11 has talked about combining two not-spectacular skills (e.g. programming and Japanese, or programming and marketing) into a very attractive package. E.g. HCI could fit with web programming, as could an appropriate hobby, as could business skills (e.g. http://www.kalzumeus.com/2011/10/28/dont-call-yourself-a-pro...).

If you just want to do something hard, AI, AR/VR, and Robotics can all work - but those are all fields which employ far fewer people than web development, and job opportunities will similarly be more limited (e.g. geographically dispersed). You'll need a bit more direction and motivation than "more defensible to future automation". To get a feel for what you like, hobby projects and e.g. Coursera can both work fine.

Alternatively, if you have a Silicon Valley salary, you could simply aim to retire by 40.

This actually was super useful, and yes, HCI is my biggest interest, but I'm having a tough time integrating it other than trying my best to be a UX expert where I work - though to be honest HCI seems to encompass a bit more than UX and it's hard to integrate the more academic aspects of HCI into simple web development.

Just being a web developer/UX expert - if you are really good at both - is already a pretty defensible position, although perhaps not a very rare one. Leveling up in HCI will help.

Another part to consider for your career strategy: if you can convince your boss that it's in his/her best interest to send you to a course or conference in topic X, that will give you both more knowledge and a semi-official "local expert" status. Once you're recognized as the local expert at X, people will bring all the hard/educational X problems to you, letting you level up quickly.

(This works best if the topic comes up quite often, but not so often that everyone has an opinion. Every web dev has some basic UX skills, but consider e.g. Information Architecture, A/B Testing, etc. - or some kind of clear "needs more-than-basic UX" category. Someone who actually knows his HCI would be in a better place to help you define the latter category, I'm afraid.)

I work in compliance technology, and I suggest you move into this area if you would like to keep programming for many more decades.

Benjamin Franklin's saying regarding death and taxes is still relevant despite being written in 1789. I believe it is also relevant in regards to creation of laws.

Last year in the US, the federal register grew by over 60,000 pages. That does not include any rules implemented by other regulatory bodies. That is just one year.

The Dodd Frank law was over 2,000 pages. The Affordable Care Act was countless thousands of pages that did not include the thousands more of rule making that would follow.

Compliance is expensive in any industry, and you have to be in compliance with the rules and regulations of your industry if you want to continue doing business.

There will always be a demand for technology related to compliance because there will always be new rules and regulations being created.

I work for the government making programs to help people comply with our regulations and we are getting to the point where the reg writers can write regs faster than we can make user interfaces for them (+ support all the existing stuff we have). Difficult to sell management on more reg writers implying a need to hire more programmers, at least until someone sues us for failing to comply with our own regulations...

There is no doubt that compliance is fertile, if boring.

I think it can also be exciting. In some cases the volume of requests requires you to research a better set of solutions.

I am given some leeway to come up with creative solutions that can scale.

Also since compliance is so broad, you get to work on many different aspects within a company. You can really improve the lives of the people in the company or institution by just coming up with an automated way to handle some of the mundane tasks involved in compliance.

I second this comment. My business (nutrition facts labeling), is totally dependent on US federal government regulations. For example, the government will be mandating new nutrition fact label designs for all food companies which is going to help open up my market greatly when the final rulings are issued.

Beyond the specifics on how websites are actually implemented, what's most important are the "soft skills", aka the people skills. Tech moves forward all the time but sitting down and interacting with a client to grok exactly what they're looking for in order to deliver ahead of schedule and below budget is what will make you stand out in a crowded space.

No. CRUD apps will always need people to implement custom business logic. Businesses will always need programmers to sit in on meetings with the client to explain what is actually going on in the code. Webservices will always be a hodgepodge mess.

In no way will this be going away anytime soon. Even if you have maybe something like factory methods to churn out your standard CRUD routes 90% of the time you've got custom logic on other routes and things like that. There will be a need for a very long time I believe.

The "Yes and No" answer by MalcolmDiggs makes good points, and I want to add to the "Yes". There has been a lot of effort around front-end design patterns that makes Web Development a lot easier. There will always be room at the top for talented individuals doing cutting edge work, but the middle and bottom end of the talent pool is being carved out by tools like bootstrap, angular, etc. You have a lot of back-end engineers doing basic data-binding on HTML, and using bootstrap to "make" the right "components", then you have a designer come in and make the CSS, etc look nice.

Add to the growing number of tools that automate a lot of the core constructs that websites now adhere, to, and...yeah...I'd say the future is kind of bleak for people who can only do JS/CSS/HTML.

<excursion> The web is starting to look a lot more standardized these days than it used to. In a lot of ways it's less impressive from a certain type of aesthetic. I remember the web circa 2009 was a lot flashier, people showed off quite a bit more than they do now. Now, users expect you to follow well established design patterns. They want to get in, get their information/data/service and get out, they don't care about your cool little reflection animation on your buttons.

In my opinion the web has kind of grown into what it's supposed to be, a fairly simple user interface around information, dynamic content, audio/video, and web applications. All of these things used to blur into one another. I think you're seeing a fundamental shift in both a labor market (web development), but also in an increasing consensus among users, and institutions as to what the web should look like. </excursion>

Edit:

I also want to add that this is what automation always tends to look like. There will always be special use cases, etc, but someone comes along and figures out how to build tools that cover 90% of people's use cases, we all adopt it and call it "standard" (cf photoshop, etc). I don't know what the back-end construct will look like, but it's happening. AWS killed a lot of potential operations work. Lambda, Heroku do a good job of abstracting 90% of most people's networking/scaling concerns. This stuff happens imperceptibly.

JS will be valuable for a long time. It's the core of the web and the web will be around for a long time, if not forever.

I find it the opposite, for every process I automate, other people are losing jobs. My code is replacing manual people processes. Someone needs to maintain that automation.