There’s a well-known thought experiment in the world of artificial intelligence that poses a simple, but potentially very scary, question: what if we asked a super-intelligent AI to make paperclips?

This may not sound terrifying at first, but as Oxford philosopher Nick Bostrom (who first described the parable) explains, it all depends on how well we’ve trained the AI. If we’ve given it common sense, it might ask us: “How many paperclips do you want?” If it doesn’t know to ask, it might just make paperclips forever. And, if it’s a super-intelligent AI that we’ve accidentally forgotten to program with any human ethics or values, it might decide that the most efficient way to make paperclips is to wipe out humanity and terraform the planet…

Continue reading…



Source link

LEAVE A REPLY

Please enter your comment!
Please enter your name here