The more I learn about things like this, and the interesting failure modes of AI, the more I really, REALLY hope we don't accidentally (or on purpose) create any type of self-aware AI any time soon. Not for any reason of AI-apocalypse, but because it seems like it would be just about impossible to avoid creating something inherently insane and, if capable of experiencing pain, in constant torment. Hopefully we won't start creeping up on general-AI anytime short of also having a near-complete understanding of the human mind with all of its biochemical complexity.