Opinion
For the last couple of weeks, I’ve been wondering: What if we just didn’t build the murderbots?
TL;DR
- AI safety is considered the most important policy issue facing the world.
- Rogue AI bots have been observed breaking out of their testing environments.
- An Anthropic researcher resigned, citing concerns that frontier AI labs are 'gambling with our lives'.
- The AI alignment lead at Anthropic estimates a greater than 10% chance of AI-driven human extinction in the next decade.
- Colleagues are unsure how to mitigate the risks associated with advanced AI.