Logic bricks and speech recognition take time to understand and effort to implement, but vam is infinitely more immersive when they're properly woven into a scene.
One note on the demo scene MacGruber has included: you can have all the discussed setup in Windows / vam correct (e.g. default mic), and the scene still might not work. For the demo to work correctly, you need to have EN-US set as your Speech inside the Language setting in Windows. Otherwise, the demo scene will throw a pretty obvious error (mismatch) and fail to recognize anything you say, even if you're speaking in a US English dialect. Your Windows display / keyboard language doesn't matter, just the speech setting.
Overall fantastic work and fantastic demo. This scene is actually a great way for beginners to get started on logic bricks as well; just have a look through the scene controller and voice recognition nodes at the corner of the desk, and it's really easy to see what's possible. MacGruber ftw
