Beliefs are Chosen to Serve Goals — LessWrong

Or: An anti-orthogonality thesis based on selection • Written as part of the MATS 9.1 extension program, mentored by Richard Ngo[1]. …

x Beliefs are Chosen to Serve Goals — LessWrong Orthogonality Thesis Agent Foundations MATS Program AI Frontpage 28 Beliefs are Chosen to Serve Goals by Ashe Vazquez Nuñez 7th Apr 2026 Linkpost for tuesdaybornwhale.substack.com 5 min read 14 28 Or: An anti-orthogonality thesis based on selection Written as part of the MATS 9.1 extension program, mentored by Richard Ngo [1] . Introduction One of the historical motivations for taking the AI alignment problem seriously is the orthogonality thesis , which states [2] : Intelligence and final goals are orthogonal: more or less any level of intellige

saved by

• Yudhister Joel Kumar
• Will Anderson
• Arya Pasumarthi

related reading

• Dialogue: Is there a Natural Abstraction of Good? — LessWronglesswrong.com
• The behavioral selection model for predicting AI motivations — LessWronglesswrong.com
• Alignment remains a hard, unsolved problem — LessWronglesswrong.com
• The hot mess theory of AI misalignment: More intelligent agents behave less coherently | Jascha’s blogsohl-dickstein.github.io
• From personas to intentions: towards a science of motivations for AI models — LessWronglesswrong.com
• Orthogonality Thesis — LessWronglesswrong.com
• why assume AGIs will optimize for fixed goals? — LessWronglesswrong.com
• Inner Alignment: Explain like I'm 12 Edition — LessWronglesswrong.com
• Optimality is the tiger, and agents are its teeth — LessWronglesswrong.com
• Towards a scale-free theory of intelligent agency — AI Alignment Forumalignmentforum.org
• AI Goals Forecast — AI 2027ai-2027.com
• Commentary on AGI Safety from First Principles — AI Alignment Forumalignmentforum.org