Deceptive AI and Society

IF 2.1 4区工程技术 Q3 ENGINEERING, ELECTRICAL & ELECTRONIC

IEEE Technology and Society Magazine Pub Date : 2023-12-01 DOI:10.1109/MTS.2023.3340232

Ştefan Sarkadi

引用次数: 0

Abstract

Deceptive artificial intelligence (AI) is a heavily loaded term. Its semantic load has become exponentially heavier in a very short period of time. Perhaps, most of this semantic load, at least in the recent public sphere, has been placed on it because of the deployment of large language models (LLMs), such as ChatGPT. Deceptive AI is very multifaceted. Different AI approaches give rise to different types of AI technologies, or, in some cases, autonomous agents. Some of these technologies already exist in practice, others exist in theory, some are transitioning between theory to implementation, and, finally, some are still only fictions of our shared imagination [62] .

查看原文本刊更多论文

欺骗性人工智能与社会

欺骗性人工智能（AI）是一个负载很重的术语。在很短的时间内，它的语义负荷已呈指数级增长。也许，至少在最近的公共领域，这种语义负担的大部分都是由于大型语言模型（LLM）的部署而造成的，比如 ChatGPT。欺骗性人工智能涉及多个方面。不同的人工智能方法会产生不同类型的人工智能技术，或在某些情况下产生自主代理。这些技术有的已经在实践中存在，有的还存在于理论中，有的正在从理论向实现过渡，最后，有的还只是我们共同想象中的虚构[62]。

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

IEEE Technology and Society Magazine 工程技术-工程：电子与电气

CiteScore

3.00

自引率

13.60%

发文量

审稿时长

>12 weeks

期刊介绍： IEEE Technology and Society Magazine invites feature articles (refereed), special articles, and commentaries on topics within the scope of the IEEE Society on Social Implications of Technology, in the broad areas of social implications of electrotechnology, history of electrotechnology, and engineering ethics.