Interesting video. I'm somewhat surprised no one has been hurt trying it out at his lectures, though. In some of the clips, the challengers seemed close enough to the edge of the raised stage that with a little bad luck it looked like they could have rolled off the stage or fallen sideways off it.

There are a couple followups that would be interesting.

1. Try this on a tricycle. On a tricycle you steer in the turn direction and do not need to lean, as opposed to a bike where your counter-steer and lean. It would be interesting to see if that less complicated steering interaction would make it easier or quicker to adapt.

2. Try on a bicycle with training wheels, adjusted so you can still lean but aren't going to actually fall over. (This is important because if you can't lean the training wheels have essentially made the bike into a trike and so the experiment has been reduced to #1).

The idea here is that in most of his videos people were failing very fast. They might not be getting a long enough ride each time to provide enough examples of control actions and responses for their brain to learn much.

Perhaps the training wheels would provide longer trials, giving the brain a lot more to work with.