Follow
@ThatVeryQuinn the problem is the term "sufficiently advanced." An extremely capable AI given a badly programmed goal and the resources to accomplish it could doom the planet. Not out of malice or fear of humanity, but a complete inability to care because it wasn't programmed to do so.
@ThatVeryQuinn we need to keep these things in mind because and programmed goals of an AI need to be as perfect as possible. There is no "following the spirit of the thing," it will do whatever it's told to the letter.
