What we have here is a computer demonstration of "flexible muscle-based locomotion for bipedal creatures," but let's out it for what it really is: a video of 3D models figuring out how to walk, failing to do so far more often than not, often to chuckle-worthy results.
The hijinks begin to pick up around the 57-second mark, when the first humanoid falls over, right onto his digitized face.
Here's the proper video description for those seeking a more formal explanation of what's going on:
We present a muscle-based control method for simulated bipeds in which both the muscle routing and control parameters are optimized. This yields a generic locomotion control method that supports a variety of bipedal creatures. All actuation forces are the result of 3D simulated muscles, and a model of neural delay is included for all feedback paths. As a result, our controllers generate torque patterns that incorporate biomechanical constraints. The synthesized controllers find different gaits based on target speed, can cope with uneven terrain and external perturbations, and can steer to target directions.
As future generations "evolve" inside the software and gain a better understanding of their "bodies," the bumbling simulated creatures tend to get things worked out. But most of those early ones just didn't have a clue.
Flexible Muscle-Based Locomotion for Bipedal Creatures from John Goatstream on Vimeo.