While there is a great point in the presentation itself - the connection between our actions and rewards makes fun, it's something we can't see in the "real lif

While there is a great point in the presentation itself - the connection between our actions and rewards makes fun, it's something we can't see in the "real life", - there is an odd summary: "The core fun in games is learning under optimal conditions". Okay, I can see why almost all fun from games is fun from learning (it seems like Ralph Koster haven't tried to eat anything tasteful, but OK, fun from games is fun from learning). But it's just =>, not <=>. Why every kind of learning (under "optimal conditions") is fun? What's new or great in learning (except for the thing you're learning - something not very real in case of games)? What's so useful in learning meaningless things, like moving Pacman, so the evolution would make us to get fun from it? And why would this strange connection last when we add learning to Microsof Excel?