AI Safety Fundamentals: Alignment cover image

AI Safety via Debate

AI Safety Fundamentals: Alignment

CHAPTER

Theorem 1: The Complexity Class of Epsilon Subscripts

There exists an X such that for all Y, H of Q, X and Y is true. Alice wins if she can find X such that all responses by Bob have H of Q equal to 1. For polynomial time H, we can continue this process for any number of rounds with Alice and Bob alternating points and counterpoints. This complexity class is epsilon subscript 2P, two steps up the polynomially hierarchy. All questions are decidable by polynometric space algorithms in SIPSA 2013.

00:00
Transcript
Play full episode

Remember Everything You Learn from Podcasts

Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
App store bannerPlay store banner