본문 바로가기
  • Home

A review of the coexistence of risks and social changes caused by Artificial Intelligence

  • Legal Theory & Practice Review
  • Abbr : LTPR
  • 2021, 9(4), pp.495-524
  • Publisher : The Korea Society for Legal Theory and Practice Inc.
  • Research Area : Social Science > Law
  • Received : November 5, 2021
  • Accepted : November 25, 2021
  • Published : November 30, 2021

Kang, Seung-Kyu 1

1호서대학교 평생교육원

Candidate

ABSTRACT

When it comes to the use of AI, data bias and ethical issues are the key. Regarding the problem of singularity (technical singularity) that AI exceeds human intelligence, famous people such as Dr. Stephen Hawking and Bill Gates have already been appealing for the danger, saying, “AI brings tragedy to mankind.” However, these claims have been toned down in recent years, and many scientists seem to be thinking, It can’t be so. It’s not impossible for AI to act autonomously, but it’s unlikely that it will become mainstream. Even now, the collective intelligence that human beings and computers form a process of knowledge is called collective intelligence. However, in the future, I think that humans and computers (data) will cooperate and work. As an example, AI is considered effective not only as a proxy for doctors, but also as a support tool for judges and police officers. As such, humans have weaknesses in several areas, so I think that if AI is applied to those areas, society will change for the better. However, a problem when using AI is the bias of data during training. US police, for example, have an AI that advises officers where to go to stop crime, called predictive guards. However, in the United States, racism is deeply rooted, and arrest rates are rising in areas with a large number of blacks. In other words, if this data is used, the AI ​​will determine that there are many suspicious people in the black residential area, creating a vicious cycle of more and more black people being arrested. If there is bias in data like this, AI will also make biased judgments, so how to exclude it is important. Not only these problems, but also numerous legal problems will arise when artificial intelligence becomes commonplace. In this background, this paper discusses how humans and artificial intelligence can coexist without risk.

Citation status

* References for papers published after 2023 are currently being built.