摘要
本文批评了目前流行的机器学习程序,如ChatGPT等,认为它们无法实现真正的人工智能,因为它们缺乏人类思维和语言的关键能力,如因果解释、道德判断和创造性批判。本文认为,人类思维是一种高效和优雅的系统,能够利用有限的信息生成复杂的句子和思想,并进行非事实和可能性的推测。本文警告,如果继续沉迷于机器学习这种有缺陷的人工智能,将会损害我们的科学和伦理,并呼吁我们重新审视人工智能的目标和方法。
This paper criticizes currently popular machine learning programs,such as ChatGPT,for failing to achieve true artificial intelligence because they lack key capabilities of human thought and language,such as causal interpretation,moral judgment,and creative critique.The paper argues that the human mind is an efficient and elegant system capable of using limited information to generate complex sentences and ideas,and to make non-factual and probabilistic speculations.The paper warns that continued addiction to machine learning,a flawed form of artificial intelligence,will undermine our science and ethics,and calls for us to revisit the goals and methods of artificial intelligence.
作者
杨旭(译)
Noam Chomsky;Ian Roberts;Jeffrey Watumull;TransYang Xu(Department of Linguistics,University of Arizona,Tucson ID85721,USA;Downing College,University of Cambridge,Cambridge ID1137455,UK;Orsinit,Honolulu ID96801,USA;School of Chinese Language and Literature,Wuhan University,Wuhan 430072,Hubei,China)
出处
《长江学术》
CSSCI
2023年第4期104-108,共5页
Yangtze River Academic
关键词
机器学习
人工智能
语言
思维
因果
道德
Machine Learning
Artificial Intelligence
Language
Thought
Causal Interpretation
Ethics