The danger of advanced artificial intelligence controlling its own feedback
(Version anglaise seulement)
par Michael K. Cohen, Doctoral Candidate in Engineering, University of Oxford
Marcus Hutter, Professor of Computer Science, Australian National University
Would an advanced AI try to take over the world? If it uses a common approach called ‘reinforcement learning’, the answer is almost certainly yes.
Lire l'article complet
© La Conversation
- dimanche 23 octobre 2022