Building and inhabiting the Virtual World
I come from a professional 3D animation background where we used 3D software Maya to design the world, the character’s, the textures and finally the animation. It was all done in a highly painstaking manner, designed, modelled, skinned and animated by a number of people to output the first norwegian 3D animated TV series (1998-2000). It’s been almost twenty years since we did that series, but the process hasn’t changed dramatically since then.
Of course there is better software, higher quality renders, real time tools like Unreal Engine which are going the VR route now. But for the most part, the approach and design philosophy is still the same. And what goes with it, the incredible amount of time it takes to arrive at the finished product. High-quality 3D animation.
Take a tool like Unreal Engine, where you, after importing the models, have to painstakingly specify all the interactive rules of the world. Although quite accessible, even to a non-programmer, it’s still an arduous process, to specify all the what ifs, and don’ts of this world.
But I belive this is about to change. With the advent of A.I., and a “What You See is What You Get” WYSIWYG approach, a small team, or even a single person, will be able to create worlds unlike anything we’ve seen before.
I believe that these tools will evolve, into high-level, easily accessible and intutive front-ends. And this will make it possible to build enourmous worlds without an army of highly specialised people.
Take a look at the evolution of HTML design. WordPress is one example where the focus, and necessity of coding, has been lifted away by such front-ends as the Visual Composer or Divi Builder. Both of them allow a WYSIWYG approach to HTML-design. This allows a speedy process, without any guesswork, of the design for the website. And if you want complexity, you can at any time dig deeper into the code, in any of the modules of these front-ends.
One can argue that the process of designing a 3D world, building the models with artistic integrity and texturing them with flare, rigging and skinning the models for animation, is so much more complex of a process. And yes, it certainly is, but this is where Artificial Intelligence algorithms will do the heavy lifting in the very near future.
These agents, which will be under the hood of every sophisticated software, will make possible a new form of communciation between the storyteller, and the tools creating the world.
Imagine the real world, now. A dialog between the production designer and the Film Director on their next project. The Film Director brings to the table the characters, the story, and how to tell it visually. While the Production Designer does everything possible to support this story, and the characters’ psychology, through the space they inhabit. Both of them have a back and forth dialog, where often the Production Designer, brings some moodboards, sketches and inspirations. Ideally the Film Director will give feedback to the Production Designer, what works, what doesn’t, and why. When they have homed in on the final design, the Production Designer will manage a team of people to realise that vision.
And now replace the Production Designer with an Artificial Intelligence agent, akin to Apple’s Siri, but inherently way more sophisticated. This agent will take propositions from the Director, or suggest something of its own accord. And then it will adjust the models, the textures and the 3d characters, all in real time, based on the Director’s input. Fine level control will always be possible, but I think the tools will get there, that it will be lightning fast, to change the physique of the characters and the appearance of the surroundings.
Change the character’s skin color, make her just a little bit older, give her more freckles on that side of the face, give her a summer dress and a hat. Now take this character and put her in this scene of where she been in a mild car crash. Rip up her clothes, give her scars, here and there, and so on and on. All of this done in real time, in a conversational manner, in VR space.
Ultimately, I am not saying that the modeler, the texturer and all these noble, and highly specialized jobs will become superfluous. Far from it, but with the use of these highly evolved A.I. agents, the coarse part of the job might be both greatly simplified, and accelerated.