Networking, analytics, multiplayer servers, etc. as a service is increasingly big business in games.

Amazon released their Lumberjack 3D engine alongside their GameLift server architecture and scaling platform: https://aws.amazon.com/gamelift/

Software Sandbox has their EpicData platform: http://swsb.io/

There's not a lot of value in making 3D engines these days, unless you're working on a real niche experience that needs something custom, a one-off.

There could be a lot of value in working on input devices, like the Leap, or Gest. But that requires really deep, low-level electrical and mechanical engineering expertise, plus probably ergonomics, physiology, machine learning, simulation, and who knows what else. You have to not only have the vision of a future input device, but a long runway to even get to prototypes, and you're behind Leap, Gest, the PlayStation Move, the Vive controllers, and Oculus Touch.

But if you want to be making the actual VR experiences, I'm not sure there's a way to do that without actually making them. All the companies in the space right now are cranking away on content and figuring things out as they go, and for all of them it's sometimes hard to tell what's a principle or maxim and what's just a current generation-hardware limitation. No-one's publishing original research on physiology and visual perception; everyone's too busy doing.

There's nothing of substance to read. There are no worthwhile classes to take. There are no projects of value to dissect.

I believe the only way forward today is to spend $1-3k on a high-end Windows 10 computer: http://arstechnica.com/gadgets/2016/03/ars-system-guide-vr-e... , buy the Rift ($600), the Vive ($800), or the HoloLens ($3000), and start learning Unity or Unreal until your VR headset shows up.

What do you build? I have no idea. What do you want to do?

360 degree video is much more open right now than games VR is, per some talks at GDC 2016 (going on right now). That requires a wide variety of skills, many more than traditional videography or filmmaking, though, skills closer to theatre production and improv and perhaps even ARGs.

The UI talks at GDC which emphasized the need to understand physiology and attention and visual fields and focus as a first-order design principle (not that it ever wasn't, but in VR it makes people sick if you don't) means that, while in server command-line UIs you could get away with an awful or arcane interface, in VR I suspect there will be a quick winnowing of any products without substantial design effort exerted on them. As a back-end developer, those probably aren't skills you have right now.

Perhaps the one saving grace everyone new to the field has right now is the quote from one of the GDC talks, "It's very possible that you might be the first person to encounter and solve a UI problem in VR."

But, VR right now isn't the web in 1999. Back then, you had HTML, you could copy and paste text and something would appear. VR right now is the design equivalent of the Homebrew Computer Club. Yes, use a 3D engine like Woz's Apple I schematics and cobble something together, but it's not going to do anything, it's not going to be anything unless you have a vision for the experiences you are working towards.

And then that experience can still only be shared by a tiny minority of people who also have $1-3k Windows 10 computers plus an additional $6-800 of VR hardware, or you have to get through Sony's licensing procedures to publish for PlayStation VR (good luck with that).

In 2-3 years the field will be completely different, the next generation of hardware will be out, and there will be paths forward, classes to take, books to read. But VR today isn't a software problem, it's a hardware problem and a content problem and a design problem.

If you don't have something you actually want to bring to life, I'm not sure it's worth spending the money. You might want to wait.