1 paper · 1 filter
Emilio Barkett, Alexander Kimpton, Daniel Graham +1
Work on the risks of artificial intelligence has focused predominantly on capability risk: the danger that systems become too powerful, too autonomous, or too misaligned with human…