Recently watching the progress made in game engines, with GI and blurry reflections, it became apparent we are rapidly approaching the point of sufficient 3d-graphics, when it’s impossible to amuse any further. I still remember the moment of characters without and then with fingers and what an impact it made. Graphics evolution was and for the time being is a huge driven force of people being interested in 3d-graphics. Let’s face it: it’s a huge impact to which game has a better graphics gamers are interested. Now imagine they are all as realistic as our cgtalk renders. What’s the point of releasing new racing game then, if a year ago it’s of the same level of realism?
And yet this leads to another issue: when the limit of textures and polygons is not a limit anymore, you don’t have to optimise as much and can use hi-poly models. One hi-poly model from turbosquid of a tank, not 100 optimised variants, which gave jobs to all those texturers and modelers making them. Only the unique content will require 3d-specialists.
And why not to use generators then? It’s perfectly possible to create generators of cities, characters and landscapes with a more or less “finite” quality. I’m sure some people already saw those, which already exist. http://www.allegorithmic.com/products/substance/designer and I think people can find city generators, landscape and character ones. Unfortunately I forgot the names of programs, but the results they produce are good enough to use. Just not for today computers.
The line between games and VFX graphics becomes slimmer and it may affect this area as well.
Just look at archviz: now people are making 10 renders instead of one, and imagine it becomes realtime what happens. Some jobs may become way less needed. This competition may affect many things.
What I see is in 10 years there will be 3d-jobs which don’t exist today, and some of today disappear. I’m also afraid the competition will become even more severe as artists will be more and more replaced with smart generators and perhaps even AI of some kind. Another possibility we will work with those generators, but creating 10 times more content than today, and perhaps for less money, as more and more people will be able to operate those programs.