Saved in:
Bibliographic Details
Main Author: Kiefer, Alex B.
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2508.20465
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912557669810176
author Kiefer, Alex B.
author_facet Kiefer, Alex B.
contents I consider motivation and value-alignment in AI systems from the perspective of (constrained) entropy maximization. Though the structures encoding knowledge in any physical system can be understood as energetic constraints, only living agents harness entropy in the endogenous generation of actions. I argue that this exploitation of "mortal" or thermodynamic computation, in which cognitive and physical dynamics are inseparable, is of the essence of desire, motivation, and value, while the lack of true endogenous motivation in simulated "agents" predicts pathologies like reward hacking.
format Preprint
id arxiv_https___arxiv_org_abs_2508_20465
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle On the possibility of deep alignment
Kiefer, Alex B.
Neurons and Cognition
I consider motivation and value-alignment in AI systems from the perspective of (constrained) entropy maximization. Though the structures encoding knowledge in any physical system can be understood as energetic constraints, only living agents harness entropy in the endogenous generation of actions. I argue that this exploitation of "mortal" or thermodynamic computation, in which cognitive and physical dynamics are inseparable, is of the essence of desire, motivation, and value, while the lack of true endogenous motivation in simulated "agents" predicts pathologies like reward hacking.
title On the possibility of deep alignment
topic Neurons and Cognition
url https://arxiv.org/abs/2508.20465