The future web developer

Bård Hovde
FED || Dead
5 min readJul 1, 2015

--

A crazy attempt at predicting the unpredictable

The other night i couldn’t sleep and did what i usually do, think of something really complex that just knocks me out.

This time though, i didn’t think about the usual stuff like “where is the universe?” or “are we all in a simulation?”, but instead “where is the web going?”

I like thought experiments, so let’s do one now. What will the job of a web developer look like in 10–20 years?

Now

Today, we hand-code websites and applications. There’s several services that can do part of that job for you — even fully automated ones — but at the end of the day, if you want something bespoke you’ve gotta build it yourself.

There’s loads of languages to write code in, frameworks to build with and libraries to help you solve challenges, and these are multiplying rapidly.

Google is already pretty smart. When I’m on my phone i rarely make it to a website in the browser any more. Usually Google will dish up the relevant information long before i get to that point, either via the search app, via Google Now, or through a native app if installed.

Social media is also swallowing up a chunk of the web. With Facebook Instant for example, Facebook says there’s even less of a reason to ever leave the app:

Leveraging the same technology used to display photos and videos quickly in the Facebook app, articles load instantly, as much as 10 times faster than the standard mobile web.

Is the “standard mobile web” doomed?

The Near Future

Let’s start with 10 years from now, 2025. Things can’t change that much in 10 years, right?

10 years ago, flash and CSS rollovers were all the rage. 2005 also gave us YouTube, Reddit and the Blaster computer worm.

First of all, the web won’t be flat.

I bet 3D modelers will be in high demand. There’s 3D printing, webGL, the still outstanding VR revolution and undoubtedly some kind of wearable tech. I can imagine 3D interface designers becoming a job title if not already.

What is the web browser used for at this point? Undoubtedly it will be used for finding information in some way, at least the information google and Facebook hasn’t yet gobbled up.

What about mobile? Is there still a mobile web?

I’m fairly certain the browser will make a comeback long before then. With Device API’s, service workers and web assembly in production, websites could perform like native apps do now.

Installable web apps is something I only just heard about and sounds very interesting. Check out the link below for a summary by @andreasbovens and @brucel.

Further into the Future

It’s the year 2035. Clint Eastwood just celebrated his 105th birthday.

What will we be building? What interfaces will people use? Some kind of contact lens? Holo-gear? Is it so ingrained there’s no UI at all? This is where things get a little blurry.

I think it’s safe to say things will be very different.

Arthur kept it old school

Maybe the web browser as we know it will still be used, but for the “Old web”. When developing for whatever interface we’ll use then, maybe supporting traditional web browsers for legacy users will feel like supporting IE6 does today?

Who does the coding though? Do we still hand write code with the loops and the if statements?

A brief disclaimer. I recently listened to an episode of The Partially Examined Life on AI that blew my mind.

It’s not hard imagining that AI eventually will take over most, if not all of the things we do today. Automation seems inevitable in this industry as most others. What better suited to write perfect code than some really clever AI?

Never tired, not reliant on caffeine, and always updated with the newest knowledge. Able to simulate every possible approach and find the perfect algorithm.

What’s hard is imagining the interface to this AI. To solve a problem, the AI needs to have a clear representation of the world and it’s environment. It needs to have goals and it needs to be flexible in the ways it can pursuit those goals.

To put it in more familiar terms.

  • The current state (a picture of the world now).
  • The target state (a picture of the world you want to end up with).
  • The intelligence to move the current state towards the target state until the two match.

This is a very general approach that is used in a lot of AI systems today, except the world that they’re modelling is very limited, and they’re used for things like showing you ads of that thing you almost bought on Amazon once when you were drunk and is now haunting you across the web.

Of course, super-intelligent AI like this is still a while away, and when we get there, building websites and apps might not even be a thing any more.

It still seems likely to me that long before then there will be some kind of software that will write software for us.

Craig tried really hard, but never seemed to get on with the humans

Conclusion

Will Google, Twitter and Facebook devour the internet? Will robots take our jobs? Will Amazon ever stop nagging about that thing?

I don’t know. Does it even matter? It’s all in a simulation anyway..

I’m tired now!

--

--