Posted 3/31/2024, 11:58:00 PM
Musk Warns of AI Dangers But Sees Benefits, While Researchers Caution of Catastrophe Risk
- Elon Musk estimates a 20% chance AI could destroy humanity, but thinks benefits outweigh risks
- AI safety researcher Roman Yampolskiy says Musk is too conservative - he puts chance of AI catastrophe at 99.999999%
- Musk said uncontrolled AI is equally dangerous no matter who makes it
- Musk advocates "truth-seeking" AI that is not forced to lie
- Geoff Hinton warns that if AI becomes much smarter than humans, it could manipulate us because it learned that from us