The AI attribution gap: Encouraging transparent acknowledgment in the age of AI

IF 3.3 2区心理学 Q1 PSYCHOLOGY, MULTIDISCIPLINARY

Intelligence Pub Date : 2025-01-01 DOI:10.1016/j.intell.2024.101880

Gilles E. Gignac

{"title":"The AI attribution gap: Encouraging transparent acknowledgment in the age of AI","authors":"Gilles E. Gignac","doi":"10.1016/j.intell.2024.101880","DOIUrl":null,"url":null,"abstract":"<div><div>The integration of Artificial Intelligence (AI), including large language models (LLMs) like ChatGPT, Claude, Gemini, and Mistral, along with specialized tools such as Google DeepMind's AlphaFold 3, is transforming the scientific discovery process. These advancements raise questions about attribution in scientific research, challenging traditional notions about the origins of discovery and the roles of human and machine collaboration. Anonymous surveys indicate that 50 to 70% of academics involved in research use AI tools. Yet, an analysis of 568 articles from three psychology Elsevier journals revealed that approximately 3.5% of these articles published since mid-2023 included an AI declaration. The reluctance of researchers to use or acknowledge AI tools can hinder scientific progress by promoting a culture wary of AI, slowing tool adoption, and limiting shared learning about their uses and limitations. Researchers are encouraged to use AI tools responsibly and detail such use in their acknowledgements to help foster a culture of transparency and innovation in scientific research.</div></div>","PeriodicalId":13862,"journal":{"name":"Intelligence","volume":"108 ","pages":"Article 101880"},"PeriodicalIF":3.3000,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Intelligence","FirstCategoryId":"102","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0160289624000746","RegionNum":2,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"PSYCHOLOGY, MULTIDISCIPLINARY","Score":null,"Total":0}

引用次数: 0

Abstract

The integration of Artificial Intelligence (AI), including large language models (LLMs) like ChatGPT, Claude, Gemini, and Mistral, along with specialized tools such as Google DeepMind's AlphaFold 3, is transforming the scientific discovery process. These advancements raise questions about attribution in scientific research, challenging traditional notions about the origins of discovery and the roles of human and machine collaboration. Anonymous surveys indicate that 50 to 70% of academics involved in research use AI tools. Yet, an analysis of 568 articles from three psychology Elsevier journals revealed that approximately 3.5% of these articles published since mid-2023 included an AI declaration. The reluctance of researchers to use or acknowledge AI tools can hinder scientific progress by promoting a culture wary of AI, slowing tool adoption, and limiting shared learning about their uses and limitations. Researchers are encouraged to use AI tools responsibly and detail such use in their acknowledgements to help foster a culture of transparency and innovation in scientific research.

查看原文本刊更多论文

人工智能的归属差距：在人工智能时代鼓励透明认可

人工智能（AI），包括 ChatGPT、Claude、Gemini 和 Mistral 等大型语言模型（LLM），以及谷歌 DeepMind 的 AlphaFold 3 等专业工具的整合，正在改变科学发现的过程。这些进步提出了科学研究中的归属问题，挑战了关于发现的起源以及人类和机器协作作用的传统观念。匿名调查显示，50% 到 70% 参与研究的学者使用人工智能工具。然而，对 Elsevier 三家心理学期刊的 568 篇文章进行分析后发现，自 2023 年中期以来发表的这些文章中，约有 3.5% 包含人工智能声明。研究人员不愿使用或承认人工智能工具，会助长一种对人工智能持谨慎态度的文化，减缓工具的采用速度，并限制对其用途和局限性的共同学习，从而阻碍科学进步。我们鼓励研究人员负责任地使用人工智能工具，并在致谢中详细说明使用情况，以帮助在科学研究中培养透明和创新的文化。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Intelligence PSYCHOLOGY, MULTIDISCIPLINARY-

CiteScore

5.80

自引率

13.30%

发文量

审稿时长

69 days

期刊介绍： This unique journal in psychology is devoted to publishing original research and theoretical studies and review papers that substantially contribute to the understanding of intelligence. It provides a new source of significant papers in psychometrics, tests and measurement, and all other empirical and theoretical studies in intelligence and mental retardation.