Saved in:
Bibliographic Details
Main Authors: Rodemann, Julian, Arias, Esteban Garces, Luther, Christoph, Jansen, Christoph, Augustin, Thomas
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2502.14581
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916731116584960
author Rodemann, Julian
Arias, Esteban Garces
Luther, Christoph
Jansen, Christoph
Augustin, Thomas
author_facet Rodemann, Julian
Arias, Esteban Garces
Luther, Christoph
Jansen, Christoph
Augustin, Thomas
contents Empirical human-AI alignment aims to make AI systems act in line with observed human behavior. While noble in its goals, we argue that empirical alignment can inadvertently introduce statistical biases that warrant caution. This position paper thus advocates against naive empirical alignment, offering prescriptive alignment and a posteriori empirical alignment as alternatives. We substantiate our principled argument by tangible examples like human-centric decoding of language models.
format Preprint
id arxiv_https___arxiv_org_abs_2502_14581
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Statistical Case Against Empirical Human-AI Alignment
Rodemann, Julian
Arias, Esteban Garces
Luther, Christoph
Jansen, Christoph
Augustin, Thomas
Artificial Intelligence
Computation and Language
Machine Learning
Other Statistics
Empirical human-AI alignment aims to make AI systems act in line with observed human behavior. While noble in its goals, we argue that empirical alignment can inadvertently introduce statistical biases that warrant caution. This position paper thus advocates against naive empirical alignment, offering prescriptive alignment and a posteriori empirical alignment as alternatives. We substantiate our principled argument by tangible examples like human-centric decoding of language models.
title A Statistical Case Against Empirical Human-AI Alignment
topic Artificial Intelligence
Computation and Language
Machine Learning
Other Statistics
url https://arxiv.org/abs/2502.14581