In I Robot, Isaac Asimov visualises a robot who murders its boss, but it turns out it is following a kind of prime directive that no harm should come to humans. (I won't say more in case anyone reading this hasn't yet read the book, or watched the film.) Sounds as though this was rather naive of Asimov ... Is there no such prime directive in place for AI?
Thanks Jane. Not as such no. AI just isn’t that powerful yet. Most directives are to try to eliminate bias or to avoid sensitive areas. But the recent studies showed AI willing to do quite nasty things even in the pursuit of “harmless business instructions”. It would be great if the world could agree on some prime directives but the prospect of them agreeing on anything does not look high.
Unsettling piece, very well written in the midst of all the AI thunder… I don’t even dare to think what the world will look like in five year’s time
So Blade runner wasn't that far out..
In I Robot, Isaac Asimov visualises a robot who murders its boss, but it turns out it is following a kind of prime directive that no harm should come to humans. (I won't say more in case anyone reading this hasn't yet read the book, or watched the film.) Sounds as though this was rather naive of Asimov ... Is there no such prime directive in place for AI?
Thanks Jane. Not as such no. AI just isn’t that powerful yet. Most directives are to try to eliminate bias or to avoid sensitive areas. But the recent studies showed AI willing to do quite nasty things even in the pursuit of “harmless business instructions”. It would be great if the world could agree on some prime directives but the prospect of them agreeing on anything does not look high.
Brrrrr. Talk about food for thought!