I think I decided finally what I want to work on for now. Along side this AI project I am involved with at the University, I am going to make significant changes to the terrain engine that can be found in the projects section.

Firstly, I am going to improve upon the generation algorithm to create a much smoother more realistic world. The method I was using before relied almost entirely on random noise and moving verticies to make sure seems lined up. I feel like this was a pretty unrealistic and poor approach when considering the performance of it. The other thing I am going to change is as in the project description make the engine completely infinite(well near infinite because of overflow) so that you can continue to wander around in the world with no errors and new chunks are generated on the fly much like Minecraft. The third thing I want to add in would be water, and possibly a river particle engine, however this will mostly depend on if I can get improve much on the performance of the previous project, however I believe that won’t be too hard to do now that I will have more time to experiment.

After the engine is running and looking cool, if I can get my hands on a DirectX11 GPU, I want to experiment with DirectCompute and populate the world with many agents to experiment with different AI algorithms on the GPU, as it is a topic I am pretty interested in. Also the engine will allow me to mess around with not only AI but possibly physics and graphical experiments as I will always have access to a different world with differing possibilities.

Anyway should be fun experimenting with all this stuff while I look for a job, and then afterwards in my free time. Excited!