Can we bootstrap AI Safety despite being unable to even define it?

Posted by cryptohell on Thursday November 13, 2025 @06:12PM
from the arxiv.org dept.

( 2 comments | 2 points | on HN )

Re:Can we bootstrap AI Safety despite being unable to even d... by cryptohell on Thursday November 13, 2025 @06:12PM (#45921963)

Given several models, assuming only that some unknown subset is "safe", can we construct a single model as safe as that subset? This reduces obtaining a trustworthy model to a plausibly easier task.

[ Parent | on HN ]

Re:Can we bootstrap AI Safety despite being unable to even d... by conartist6 on Thursday November 20, 2025 @12:44PM (#45995367)

AI output is modeled on human behavior. Are humans safe?

[ Parent | on HN ]