CosmicSex on 03 January 2015
Being as though we would have to program it, we could probably save ourselves a lot of trouble by creating none malevolent AI. As a hobby I like to program and the first thing you realize when iterating through code is that it does exactly what you tell it to.
As long as we don't program it to destroy us all or enslave us we should be safe. The problem is the programmer, not the program. Always.
