I read an article by Michael Abrash on the Valve blog that he signed up with Valve to investigate Augmented Reality. At the end of his article he mentions that Valve is hiring.
This prompted me to send him the following long email. And as I like for my writing to not go to waste I am posting it here as well.
I read the Valve employee handbook directly followed by your great article. There are many concepts in Valve’s philosophy that resonate with me. However I have the same problem that most people who read this have: Geography.
Granted, anybody is – theoretically – free to move to live near Valve’s Headquarters and few people do… The handbook mentioned like 44 international hires so far. This number is so low because of several obvious reasons
- too few H1B visas
- most engineering jobs in the USA require a bachelor’s degree or higher
- … Because that’s needed for the visa
- … But in many countries the education of engineers is vocational and not academic and therefore people have no such academic degree and it is quite a fuss to formally prove that the education and/or experience is equivalent.
- very few people are actually willing to move to the USA because the the quality of living is far greater elsewhere
- … Or they are married (like myself) and there is no chance our wives would ever consent to moving away from our rural town, let alone to the USA
But I am not writing to you because I am whining. I want to open your eyes to a problem that technology companies in the USA are facing. And where there is a problem an engineer will work on a solution.
To give an analog from recent history: if people ship physical goods the recipient has to pay customs taxes. This is why a recent Kickstarter project has a rewards tier that is entirely virtual. You get everything in digital form and downloadable over the internet as to not having to pay any customs or taxes for it.
Which brings me to the point I want to address: somebody needs to invent a way that somebody in a foreign country could work at Valve virtually. Basically a level of cyberspace that is of sufficient fidelity that it can replace actual in-person presence. There’s another Kickstarter project that went into this direction. It was a robotic cradle for an iPhone that you could remote control. The need to see and hear from a remote location, enhanced by the need that you want to be able to change the perspective.
Granted there are some ways how companies try to address this: Skype, Go2Meeting, video chat, Sharepoint. The most innovative development probably being Microsoft’s RoundTable device. But they all fall short because they simply don’t have the multi-disciplinary makeup of researchers that you need to solve this hard problem.
3D-Graphics and Immersion and virtual worlds and internetworking have always been the forté of Game companies like Valve. Companies like Google or Microsoft are working on small slices of the problem as they understand it. But where is the enthusiasm for immersive cyberspace that we all felt 20 years ago?
3D TV has become an “also ran”, cyperspace is no longer of interest for anybody, even Google with their glasses project is only focussing on displaying information (and probably enhance our reality with ads).
The great opportunity for Valve is to enhance the Source engine in several key areas:
- revolutionise motion capture and eliminate the need for markers
- implement real time 3D tracking and “world building”
- develop real time interpolation between these 3D scans
- let multiple such live 3D scanners work in tandem on internetworked servers
- develop mobile clients that can tap into these worlds and navigate them
- develop technology that can insert a remote worker’s avatar into your current reality
I think that the sub-problem of real time scanning and world construction can be solved sooner than the problem of how to get this onto people’s retinas. But the immediate benefits would be a dramatic increase in productivity for 3D graphics as used in computer animated and 3D live action movies. Also I think many people would be interested in being able to go into a room, taking a look around once and then having a 3D model of that for later use.
Sort of like the holo cam that Lt. Data had as a hobby in Star Trek: TNG.
Again it has to be a Gaming company to provide the technology for this next paradigm shift. Camcorder companies can only increase the number of pixels on their CCDs. And their “solution” for the 3D model problem is to not solve it, but let the viewer’s brain figure this out.
The next quantum leap in vision technology will no longer be to record pixels, but to record an actual 3D model of the real world. And I think this will come as a side-product of the greater goal of being able to “beam” reality to other places on earth for virtual consumption first and actual participation later.
Speaking of consumption … Could you please check on the progress of HL2:ep3?