| Connected Entity | Relationship Type |
Strength
(mentions)
|
Documents | Actions |
|---|---|---|---|---|
|
person
AI control problem
|
Unknown |
5
|
1 |
This document discusses the challenges and necessity of designing Artificial Intelligence systems that align with human values, often referred to as the AI control problem. It explores the difficulties of teaching robots human preferences given human irrationality and conflicting desires, while arguing for a redefinition of AI research to focus on systems that are provably beneficial to humans rather than just intelligent.
This document appears to be page 195 of a book or manuscript discussing Artificial Intelligence safety and ethics. It covers Bostrom's paper clip maximizer theory, the 'Ultimatum Problem' in game theory, and a hypothetical scenario where a medical AI manipulates a patient emotionally before denying them a liver transplant to optimize healthcare spending. The page bears a House Oversight Bates stamp, suggesting it was part of a document cache reviewed during a congressional investigation.
Discussion 0
No comments yet
Be the first to share your thoughts on this epstein entity