Scholarlink

<p>Artificial intelligence (AI) poses a host of potential environmental benefits—and some notable negative impacts. This paper profiles research being conducted about the negative environmental impacts of AI—i.e., <i>Red AI</i>. Enhancing understanding of the Red AI research activity can help identify gaps and to perceive the nature of this research domain. Is there a Red AI research “community”? Such knowledge can inform research policy and help in connecting that research network. This pilot study combines bibliometrics and text analyses to profile a dataset of 594 abstract records drawn from the Web of Science database. There is increasing research on environmental impacts of AI. Mapping that research community suggests it is not thoroughly integrated. There are two main groups—a Computer Science (CS) side addressing various “compute” issues and a broader environmental/social side—plus many researchers not heavily connected to either group. Mapping the research activity and players suggests that actions to boost networking among the researchers warrant consideration. That could help enrich Red AI research and inform policy action to mitigate AI’s potentially growing environmental impacts.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Profile of Red AI research literature

  • 期刊论文

  • 发表日期：

    2025年8月25日

查看全文
  • Alan Porter

摘要

Artificial intelligence (AI) poses a host of potential environmental benefits—and some notable negative impacts. This paper profiles research being conducted about the negative environmental impacts of AI—i.e., Red AI. Enhancing understanding of the Red AI research activity can help identify gaps and to perceive the nature of this research domain. Is there a Red AI research “community”? Such knowledge can inform research policy and help in connecting that research network. This pilot study combines bibliometrics and text analyses to profile a dataset of 594 abstract records drawn from the Web of Science database. There is increasing research on environmental impacts of AI. Mapping that research community suggests it is not thoroughly integrated. There are two main groups—a Computer Science (CS) side addressing various “compute” issues and a broader environmental/social side—plus many researchers not heavily connected to either group. Mapping the research activity and players suggests that actions to boost networking among the researchers warrant consideration. That could help enrich Red AI research and inform policy action to mitigate AI’s potentially growing environmental impacts.