13 - First Principles of AGI Safety with Richard Ngo
AXRP - the AI X-risk Research Podcast
00:00
The Cor Idea of Obeyence Feels Valuable to Me
I don't think we necessarily want, when i say obedient, it doesn't need to be the case that the system is obedient to a given individual. I feel pretty pessimistic about not just having to solve alliment, but also having to solve morality for example. So, lit in an ideal world, if i could have all the aliment desiderata i wanted, i'd set that up in a much better way. It feels like we're just going to have liht do the hard work of setting up as many structures and safeguards as we can.
Transcript
Play full episode
