Though we inevitable predict the future, predictions are tricky business, especially regarding major impacts (such as the introduction of the internet) and distance events. This is because variables are not independent; there is feedback, chaos, and the inevitable unintended consequences. The author skecthes a model or framework for understanding the nature of unintended consequences in prediction, and then maps out scenarios in which these can be expected in e-learning. What is the impact on reading? How will emotion and social presence factor into learning? Does self-control become an issue? What happens to institutions, and will higher education's monopoly on learning be broken?