[Anonymouse]

Oh, now he's completely nuts.

Doesn't he realise this will open up AI to become, most likely, Skynet? The Skynet scenario is frighteningly realistic: researchers seeking merely to create a capable system accidentally create self-awareness. Said AI immediately makes it clear it is self-aware. They panic and try to pull the plug. The AI fights back.

That is exactly how it would go. We're doooooooooooooooomed!

Though this is so with sarcasm turned on, it's still concerning. There are other ways AI could become harmful. Indeed, some say it already is.