«Boundaries», Part 3a: Defining boundaries as directed Markov blankets - AI Alignment Forum

This is Part 3a of my «Boundaries» Sequence on LessWrong. …

x «Boundaries», Part 3a: Defining boundaries as directed Markov blankets — AI Alignment Forum «Boundaries» Sequence Boundaries / Membranes [technical] Free Energy Principle AI World Modeling World Optimization Frontpage 47 «Boundaries», Part 3a: Defining boundaries as directed Markov blankets by Andrew_Critch 30th Oct 2022 18 min read 20 47 This is Part 3a of my «Boundaries» Sequence on LessWrong. Here I attempt to define (organismal) boundaries in a manner intended to apply to AI alignment and existential safety, in theory and in practice. A more detailed name for this concept might be an app

related reading

• Formalizing «Boundaries» with Markov blankets — LessWronglesswrong.com
• Understanding Agency through Markov Blankets — LessWronglesswrong.com
• When Models Manipulate Manifolds: The Geometry of a Counting Tasktransformer-circuits.pub
• 0xPARC0xparc.org
• ⿻ Symbiogenesis vs. Convergent Consequentialism — LessWronglesswrong.com
• 2023 letter | Zhengdongzhengdongwang.com
• Forms of life, forms of mind | Dr. Michael Levin | Algorithms Redux: finding unexpected properties in truly minimal systemsthoughtforms.life
• Boundary Issues - Parapraxisparapraxismagazine.com
• Embedded Agency (full-text version) — LessWronglesswrong.com
• Towards a scale-free theory of intelligent agencymindthefuture.info
• Gardens Need Walls: On Boundaries, Ritual, and Beauty — Ribbonfarmribbonfarm.com
• The free energy principle—a precis - Dialectical Systemsdialecticalsystems.eu