Superintelligence: Paths, Dangers, Strategies · chapter 8 · id superintelligence-c8-02-intelligence-and-final-goals-a
“Intelligence and final goals are independent -- a superintelligent system could in principle have any goal, including trivial ones. This is the 'orthogonality thesis.'”
holdsconfidence: high⚠ extracted by pipeline, re-audit pending
Receipts
Wikipedia: Orthogonality Thesissource alive
The orthogonality thesis states that an agent can have any combination of intelligence level and goal, with some caveats.
Hume's Is-Ought Problem (philosophical foundation)source alive
David Hume argued in 1739 that one cannot derive an 'ought' from an 'is.' Factual knowledge does not logically constrain goals -- the philosophical root of the orthogonality thesis.
Stuart Russell, Human Compatible (2019)source alive
Russell largely accepts the orthogonality thesis and builds his alignment framework around it, treating it as a design constraint for beneficial AI.
This claim is a stable, citable object. If you can falsify a verdict, tell us — corrections are loud here.