东北大学学报（社会科学版）

东北大学学报（社会科学版） ›› 2019, Vol. 21 ›› Issue (1): 14-19.DOI: 10.15936/j.cnki.1008-3758.2019.01.003

• 科技哲学研究 • 上一篇    下一篇

人工智能威胁论:逻辑考察与哲学辨析

李帅   

  1. (南开大学哲学院，天津300350)
  • 收稿日期:2018-06-20 修回日期:2018-06-20 出版日期:2019-01-25 发布日期:2019-01-12
  • 通讯作者: 李帅
  • 作者简介:李帅(1991-)，男，湖北荆州人，南开大学博士研究生，主要从事逻辑学、科学哲学研究。
  • 基金资助:
    国家社会科学基金重大资助项目(15ZDB018);教育部人文社会科学研究资助项目(18YJC72040001);天津市社会科学重点资助项目(TJZX17-001)。

Threat of Artificial Intelligence: Logical Investigation and Philosophical Analysis

LI Shuai   

  1. (College of Philosophy， Nankai University， Tianjin 300350， China)
  • Received:2018-06-20 Revised:2018-06-20 Online:2019-01-25 Published:2019-01-12
  • Contact: -
  • About author:-
  • Supported by:
    -

RichHTML

0

PDF (PC)

1306

摘要： 近来年，人工智能发展异常迅猛，引发了一些学者的担忧，他们认为人工智能在不久的将来可能会威胁人类生存。从哲学和逻辑的角度考察这个议题，试图构造一个溯因推理论证驳斥这一论题。如果人工智能会威胁人类生存，那么它必然拥有情感认知能力。获得这种能力可能要借助全脑仿真技术，而这项技术的实现过程中会遇到“指数爆炸”问题。威胁论的论证基于归纳推理，不具备必然的可靠性。但如果仅将威胁论视为一个信念，进行一场类似“帕斯卡赌”的赌局，考虑到信念成真的巨大代价，有必要采取一些防范措施。

关键词: 人工智能, 威胁, 溯因推理, 帕斯卡赌

Abstract: In recent years， with the rapid development of artificial intelligence， some scholars are concerned that in the near future， artificial intelligence may threaten human survival. This issue is examined from the philosophical and logical point of view so as to construct an argument of abduction. If artificial intelligence threatens human survival in the future， it must have emotional cognition. To acquire this ability， whole brain simulation is likely to be relied upon and the realization of this technology may face the “exponential explosion”. The current argumentation of this hypothesis is based on induction and does not have the necessary reliability. However， if the thesis of threat is only treated as a belief whereby a gambling similar to the “Pascal's wager” is conducted， risks of a belief becoming the reality can be huge. Thus， it is necessary to take some precautions.

Key words: artificial intelligence, threat, abduction, Pascal's wager

中图分类号: