Conversation

This is a very helpful model for thinking carefully about AI risk and safety
Quote Tweet
One of the ideas I find most useful from @AnthropicAI's Core Views on AI Safety post (anthropic.com/index/core-vie) is thinking in terms of a distribution over safety difficulty. Here's a cartoon picture I like for thinking about it:
Show this thread
How Hard Is AI Safety? A graph with three distributions over difficulty "Safety isn't a problem view" (focused on easy), "Anthropic" (widely distributed), and "Pessimistic Safety View" (focused on AI is very hard).
This is interesting. I wonder how are you practically choose projects along those curves, or indeed try to find the shape of the curve.
2

Discover more

Sourced from across Twitter