포스트휴머니즘 시대에 인간과 인공지능 이해하기: ‘강한 인공지능’의 등장에 어떻게 대처할 것인가?

소형근 한국구약학회 2022 구약논단 Vol.28 No.4

This paper introduces the artificial intelligence technologies realized in the posthumanism era, and in particular, the current state and problems of ‘Strong AI’ technology, and the purpose of human creation from a biblical perspective, and suggests the response to ‘Strong AI’. The emergence of ‘Strong AI’ is highly likely to eventually lead us to a dystopian world. The reason is that firstly humans reaches a point where ‘Strong AI’ cannot be controlled, secondly humans with a sinful nature misuse ‘Strong AI’ for private purposes, thirdly if ‘Strong AI’ itself makes a wrong judgment, like Victor, the main character of <Frankenstein>, the human being has no choice but to end in ruin. AI should exist for mankind, and if it is harmful to mankind, it will be out of the way of the purpose of production. However, it is practically impossible to delay or curb the development of ‘AI’. Humans were created as beings who work(Gen 2:15) a lower than God(Ps 8:5), and were created for the purpose of maintaining and preserving God’s creation order as God’s representatives (Gen 1:28), to glorify God the creator(Isa 43:7), and to praise God(Isa 43:21). If so, the reason for the existence of AI(or ‘Strong AI’) becomes also clear. AI should bring positive and beneficial effects to mankind. For that, the personality of a programmer is important, and spirituality that will illuminate the meaning of human beings and life is also needed. However, it is a passive regulation to rely only on individual personality or individual spirituality for AI development, and the ‘legal regulation’ of the authoritative world organization for AI development has been codified and regulated as international social agreement, and management and supervision must be accompanied. Now more than ever, communality should be a major topic for mankind. In addition to the ‘climate crisis’, ‘energy crisis’, ‘food crisis’, ‘epidemic spread crisis’, and ‘low birthrate’, and ‘aging crisis’, it is important how to use and control AI that takes the place of humans and sometimes threatens them. A joint disposal for these problems should be an urgent issue for our humanity. Humanity will have to respond to ‘Strong AI’ with a common goal as a response to these crises.

대학생의 약·강 AI 인식 집단 간의 차이 분석: 개인적, 심리적, 윤리적 쟁점을 중심으로

신나민,김종욱 한국지식정보기술학회 2021 한국지식정보기술학회 논문지 Vol.16 No.5

This study analyzed group differences between weak and strong AI perceptions over personal and psychological variables and ethical issues. The research participants were 349 university students (female=60.2%, natural science/engineering major=71.6%, strong AI group=58.2%) recruited from three 4-year universities located in the city of Seoul, Busan, and Kyung-nam province. The survey was administered online; and collected data were analyzed via spss version 21.0 using t-test, cross-tab and frequency analyses. The results of the data analysis are summarized as following. First, no statistical differences were found out between the two groups in the variables of sex, major, experience and knowledge of AI. Second, psychological variables such as negative attitude towards AI, anxiety, pessimistic and optimistic scores proved to make a difference between the two groups. Weak AI groups were more negative, anxious, pessimistic and less optimistic towards AI than strong AI groups. Third, no difference was observed in their perceptions of usefulness and easiness of AI technology. Fourth, perceived need for AI ethics education is higher in strong AI groups than weak AI groups. Fifth, among the 12 application areas of AI, both groups ranked high in service (1st) and transportation (3rd) sectors along with translation and medical fields ranked second in each group. Sixth, regarding the subject of AI ethics, both groups gave developers strong responsibilities in the areas of self-driving car, medical services, education, elderly care, and military robots/AI. In addition, the degree of ethical requirements for manufacturers, service providers, and users were found to vary in each areas. The results of this study suggest that an individual's perception of AI has to do with psychological variables, but not directly related to the acceptance of AI technology. Regarding AI ethics, this study revealed that different weights of responsibility are given to differing AI applications and stakeholders.

확장된 계획적 행동이론을 통해 본 강한 인공지능 제품에 대한 이용자의 수용의도: 20대 연령층을 중심으로

이창섭,이현정 한국콘텐츠학회 2020 한국콘텐츠학회논문지 Vol.20 No.10

The rapid progress of AI technology gives us the expectation to solutions to various problems in our society, and at the same time, it gives us anxiety about the side effects that can occur if AI develops beyond human control. This study was conducted in the early 20s with less objection to advanced devices. We attempted to provide clues to understand thoughts and attitudes of the targets about the future environment that will be brought by AI through the process of finding intent the acceptance of strong AI technology. For this, we applied the Theory of Planned Behavior, and further expanded this research model to identify factors affecting the attitude toward AI. As a result, the attitude toward AI and perceived behavioral control had a significant effect on the intention to use to strong AI. In addition, we found that the expectation of the benefit of improving task performance and the anxiety on the threat of relationship disturbance had a significant effect on the attitude toward AI. This study suggests implications for AI-related companies establishing the direction of technology development and for government setting a policy direction for AI adoption. AI 기술의 급격한 진보는 우리 사회의 다양한 문제점을 개선해주는 기대를 주는 동시에, 인간이 제어할 수 없을 만큼 발전했을 경우 일어날 수 있는 다양한 부작용에 대한 불안을 안겨준다. 본 연구는 첨단 기기에 대한 거부감이 적은 20대 초반 젊은 층을 대상으로 강한 AI 기술이 적용된 제품의 수용의도를 알아보는 과정을 통해 AI에 의해 도래할 미래 환경에 대한 이들의 생각과 태도를 이해하는 단서를 제시하고자 하였다. 이를 위해 계획적 행동이론을 적용하였고, 나아가 이 연구모델을 확장시켜 AI에 대한 태도에 영향을 미치는 요인들을 확인해보고자 하였다. 연구 결과, AI 제품 사용의도에 AI에 대한 태도 및 지각된 행동통제가 유의한 영향을 미쳤고, AI에 대한 태도에는 업무 성과 향상의 혜택에 대한 기대와 관계 교란의 위협에 대한 불안의 영향을 확인하였다. 본 연구는 AI관련 기업 관점에서는 기술개발의 방향을 세우고, 국가적 관점에서는 AI의 정책적 수용 방향성을 세우는 데 시사점을 제시하며 연구적 공헌점을 가진다.

인공지능(AI)과 윤리

김영한(Yung Han Kim) 한국신학아카데미 2026 신학포럼 Vol.21 No.-

인공지능 기술은 인류 생활의 획기적인 발전을 가져다주었다. 오늘날 인공지능(Artificial Intelligence, AI)이 불러올 미래적 현실에 대한 유토피아적 낙관주의(utopian optimism)가 가시화되고 있으나 이와 더불어 디스토피아적 비관주의(dystopian pessimism)가 혼재되어 등장하고 있다. 더욱이 인간 지능이 단순한 보조 정보도구인 약(弱)AI를 넘어서, 인간과 대등한 수준의 강(强)AI, 그리고 인간지능을 너머선 초(超)AI 수준으로 나아가게 될 때, 인공지능은 도구적 차원에서 이탈하여 인간 존재에 대한 존엄성, 통전성, 자유성, 결정권, 사고력에 대한 침해로 이어지게 된다. 여기서 인공지능이 하나의 기계를 넘어서서 하나의 대행자(agent)가 되어 윤리의 문제를 유발시킨다. 인공지능 자체의 윤리는 인간 및 인류의 생존에 심각한 영향을 끼치게 된다. 그러므로 인공지능의 윤리 문제는 중요한 이슈가 된다. AI는 인간의 지적인 능력을 확장하는 강력한 도구로서 데이터를 기반으로 학습하고, 반복을 통해 판단하고, 심지어 창의적인 작업까지 수행하고 있다. 하지만 AI는 인간의 본질을 구성하는 자기 인식, 도덕적 책임, 그리고 영적, 감정적 차원의 깊이를 수행할 수 없다. 인간이 만든 로봇에게 윤리를 주입시키는 것은 인간의 몫(portion)이요, 이 윤리 체계는 인간 윤리의 패러다임에 달려 있다. 좋은 윤리 체계가 주입된 로봇은 선한 일을 하나, 나쁜 윤리체계가 주입된 로봇은 인간 공동체에 해와 악을 가져다 줄 수 있다. 이런 의미에서 인공 지능 사회는 인간에게 판도라 상자라고도 말할 수 있다. AI시대 지성인들은 디지털 문해력(digital literacy)를 익혀야 한다. 어떤 알고리즘이 사람을 조작하는가, 어떤 윤리적 위험이 있는가, 데이터는 중립적인가, 이런 질문들에 대해 스스로 판단할 수 있는 안목을 키워야 한다. 교회와 신앙공동체는 AI를 유용한 도구로 활용하되, 인간의 존엄성과 관계성을 중심에 두고 사회적 공동선(자유, 평화와 평등) 추구를 위한 윤리적 가이드라인을 마련해야 한다. AI-Technique has brought an epoque making development in the life of humankind. Concerning the influence of AI, an utopian optimism and at same time a dystopian pessimism have appeared in our life situation today. AI(Artificial Intelligence) is expected to go over the level of weak AI, which is a mere information instrument, and to come to the level of super-AI, which is beyond human intelligence, going over the level of strong AI, which is the level equal to humans. In that case, AI will infringe upon the human dignity, integrity, freedom, decision and thinking, parting from the level of instrument. Here, AI becomes an agent, causes the ethical problems. The ethic of AI will influence a serious impact on human existence. The ethic of AI becomes an important issue. AI is a strong instrument expanding human intelligent capacity, learns based on the data, judges through repetitions, performs even a creative things. However, AI cannot perform the self cognition, moral responsibility and spiritual, emotional depth. AI needs human portion for introducing an ethic into itself. AI’s ethical system depends upon the paradigm of human ethic. The robot into which a good ethic is put, does good things, The AI-robot a bad ethic put into can do harm and evil to the humankind. In this sense the AI society can be called the pandora box. The citizen of AI age is advised to get familiar with a digital literacy. Citizen needs to get view ability to judge such issues as which kind of algorithm manipulates the human, what kind of ethical danger exists, and whether data are neutral or not. Church and faith community need to set up the ethical guideline for searching for social common good(freedom, peace, and equality), using AI as a useful instrument, based on the human dignity and relationship.

인공지능(AI)과 형사책임 –‘형사적인 것’의 탐구를 위한 문제 제기를 겸하여 –

오병두 홍익대학교 법학연구소 2025 홍익법학 Vol.26 No.4

인공지능(AI, Artificial Intelligence) 기술의 급속한 발전과 활용영역의 지속적 확대는 형사법이론, 특히 신중한 접근이 요구되는 형사실체법 분야에서 중대한 도전 이 되고 있다. 학설은 인공지능이 점차 인간과 유사한 기능을 수행함에 따라 형법이론 을 적용하여 인공지능에 대해 형사책임을 부과할 수 있는지를 중심으로 논의가 전개되고 있다. 인공지능의 형사책임을 인정하기 위해서는 인공지능을 독자적 형사책임 귀속의 주체로 인정할 수 있어야 한다. 학설은 이를 위해서 인공지능이 야기한 외부적⋅물리 적 ‘사태’를 인간의 ‘행태’와 동일하게 법적으로 취급하고자 한다. 법기술적 차원에서 인공지능을 ‘전자인(e-person)’으로 간주하거나, ‘인간(Mensch)’과 ‘인격(Person)’을 분리하면서 후자를 중심으로 법이론을 구성하는 ‘탈인간중심적 법이론’에 근거하여 형사책임을 인정할 수 있다거나, 인공지능의 강⋅약 구분론을 토대로 강한 인공지능 에 대해서는 독자적인 행위능력을 인정할 수 있다는 등의 접근이 나오고 있다. ‘인격’ 개념을 중심으로 하는 접근은 특정한 법 영역에서 권리·의무의 귀속을 인정하기 위한 제한적 논의를 그대로 형사책임 주체성의 근거로 삼는다는 점에서 설명력이 충분하지 않다. 또한, 인공지능의 강·약을 구별하는 것은 형사책임의 최종적 귀속 관계를 해명하는 기준이 되기 어렵고 그 배후에 있는 인간의 책임을 은폐하거나 회피하는 결과가 될 여지도 있다. 따라서 이상과 같은 논의는 별도의 근거가 제시되지 않는 한, 인공지능의 형사책임을 충분히 해명하기 어렵다. 또한, 이를 전통적 형법이 론에 적용하는 경우, 인공지능에 대해 인간을 대상으로 하는 행위능력, 책임능력이나 형벌능력(수형능력)을 그대로 인정하기도 어렵다. 인공지능의 형사책임 논의는 그간 형법이론에서 자명한 것으로 간주되었던 요소 의 의미와 그 타당성을 묻고 그 토대부터 재검토할 것을 요구한다. 인공지능의 형사책 임은 규제 목적에 비추어 인간과 동일하게 모든 범죄에 적용되기보다는 제한된 영역과 특정 행위 유형(엄밀하게는 ‘사태’ 유형)에 한정하여 법률이 정하는 바에 따라 인정될 가능성이 크고, 그 금지의 범위와 제재는 형사정책적 목적과 요구에 따라 결정될 문제라는 것이 일반적인 시각이다. 그렇다면 인공지능의 형사책임 문제 는 결국 특정 위반 행위나 제재에 ‘형사적인 것’이라는 속성을 부여하는 요소가 무엇인지 규명하는, 즉 ‘형사적인 것’의 의미를 탐구하는 문제로 귀결된다. 이 도전에 답하는 것이 ―인공지능에 대한 것에 국한되지 않고― 형법이론의 과제이다. The rapid development and continuous expansion of Artificial Intelligence (AI) technology present significant challenges to criminal law theory, particularly in the area of substantive criminal law. Scholarly discourse addresses the attribution of criminal liability to AI, questioning whether current criminal law frameworks are applicable as AI increasingly performs functions analogous to those of humans. For AI to be held criminally liable, it must first be recognized as an independent subject of criminal liability. Arguments supporting AI's criminal liability include comparing AI-caused events to human actions; granting AI legal personhood (e-person) as a legal entity responsible for criminal acts based on post-human legal theories; and recognizing an independent capacity for action in strong AI, distinct from weak AI. However, concentrating solely on AI as a rights-bearing and duty-bound subject within a specific legal framework does not always resolve the issue of criminal liability. Moreover, distinguishing between strong and weak AI is an inadequate criterion for determining the ultimate attribution of criminal liability and risks obscuring or evading the accountability of the humans responsible for the AI. The discussion surrounding AI's criminal liability necessitates a critical reassessment of the fundamental concepts traditionally accepted in criminal law theory. Considering regulatory goals, the prevailing view is that AI is unlikely to be held criminally liable under the same standards applied to humans for all offenses. Instead, liability will be confined to specific areas and particular types of actions—namely, those events legally defined as crimes. The scope of prohibitions and penalties will be determined by the goals and requirements of criminal policy. This means that determining AI's criminal liability ultimately depends on identifying the elements that make a particular act criminal or punishable—that is, understanding what constitutes ‘the criminal.’ Responding to this challenge falls within the task of criminal law theory, which is not limited to AI alone.

인공지능에 대한 비판적 고찰

이상원(Sangwon Lee) 신학지남사 2017 신학지남 Vol.84 No.4

The power of logical reasoning and calculation themselves are natural technology and the expressions of the general grace of God and the general operation of the Holy Spirit. But, if the powerful technology of weak AI is abused, it can give harms to human society. Therefore it has to be under the moral control of man. The technology separated from moral directions can fall into inhuman and destructive instruments. Asimov’s three laws of a robot is one of the efforts subjecting weak AI to the moral control of man. Weak AI imitates a part of human intelligence alone, not other functions like emotion, will, conscience, free choice, and physicality. Therefore, even if its reasoning and calculating ability is excellent, it is just a fake of human personality. The reductionism of weak AI technology can dehumanize our human society. Strong AI is constructed imaginatively, when too much confidence is given to the power of weak AI. Speculative scientists and science fiction authors are certain that in the near future AI will imitate all of the functions of the human spirit. But strong AI is not the result of scientific study but a sort of myth. Strong AI is a superman. Speculative science focuses upon utopia, but science fictions upon dystopia. Too much fear of AI technology can generate too much attachment to dystopia. This fear comes from the false application of the present advanced up-to date scientific technology to the future human society without considering that future human beings will meet the same technology with more advanced spiritual and physical powers and devices. But this fear can give us serious warnings not to abuse AI technology. A too much attachment to utopia is dangerous. The attempt to get eternal life and to build up utopia with technology is spiritual arrogance before God. This attempt cannot be realized, and can lead mankind into a big disaster. Furthermore the utopia of scientific speculation and science fictions is pantheistically confined in the present society. But the Scriptural ideal society is dynamic, abundant, and personal. It has a continuity with the present world in its outer form and identity, but its content has a discontinuity with the present world in that its content is wholly new ones. This ideal society will be given by the grace of God alone.

A study on legal service of AI

Jong-Ryeol Park(박종렬),Sang-Ouk Noe(노상욱) 한국컴퓨터정보학회 2018 한국컴퓨터정보학회논문지 Vol.23 No.7

Last March, the world Go competition between AlphaGo, AI Go program developed by Google Deep Mind and professional Go player Lee Sedol has shown us that the 4th industrial revolution using AI has come close. Especially, there ar many system combined with AI hae been developing including program for researching legal information, system for expecting jurisdiction, and processing big data, there is saying that even AI legal person is ready for its appearance. As legal field is mostly based on text-based document, such characteristic makes it easier to adopt artificial intelligence technology. When a legal person receives a case, the first thing to do is searching for legal information and judical precedent, which is the one of the strength of AI. It is very difficult for a human being to utilize a flow of legal knowledge and figures by analyzing them but for AI, this is nothing but a simple job. The ability of AI searching for regulation, precedent, and literature related to legal issue is way over our expectation. AI is evaluated to be able to review 1 billion pages of legal document per second and many people agree that lot of legal job will be replaced by AI. Along with development of AI service, legal service is becoming more advanced and if it devotes to ethical solving of legal issues, which is the final goal, not only the legal field but also it will help to gain nation’s trust. If nations start to trust the legal service, it would never be completely replaced by AI. What is more, if it keeps offering advanced, ethical, and quick legal service, value of law devoting to the society will increase and finally, will make contribution to the nation. In this time where we have to compete with AI, we should try hard to increase value of traditional legal service provided by human. In the future, priority of good legal person will be his/her ability to use AI. The only field left to human will be understanding and recovering emotion of human caused by legal problem, which cannot be done by AI’s controlling function. Then, what would be the attitude of legal people in this period? It would be to learn the new technology and applying in the field rather than going against it, this will be the way to survive in this new AI period.

AI 기반 강군 건설을 위한 전장 데이터 구축-연결-보호 방안

이재숭,정회진,최영태,김강용 한국통신학회 2026 한국통신학회논문지 Vol.51 No.1

The accelerating development of artificial intelligence (AI), coupled with recent experiences in ongoing conflicts, underscores the extent to which emerging technologies are reshaping the fundamental paradigm of warfare. Confronted with structural challenges such as declining birth rate and the transformation of warfighting paradigms, the ROKA has launched Defense Innovation 4.0. This study examines strategies for constructing a Strong Military through AI-driven Science and Technology and critically analyzes the key challenges encountered in this transformation. By contextualizing the U.S. OFFSET strategies within the broader trajectory for the ROKA's pursuit of Army TIGER, an AI-driven future force initiative. The research highlights three fundamental challenges in advancing AI-enabled defense innovation: the systematic development of battlefield data, the integration of hyper-connected data, and the enhancement of data-centric security. The study further explores potential approaches to mitigate these challenges. Ultimately, the construction of a Strong Military through AI-driven Science and Technology requires more than technological innovation; it necessitates a comprehensive paradigm shift encompassing large-scale data governance, resilient connectivity, and robust data protection mechanism.

인공지능(AI)과 지능형 로봇의 법인격에 대한 소고

정영진(Young-Jin Jung) 4차산업혁명융합법학회 2021 4차산업혁명 법과 정책 Vol.4 No.-

2012년 이미지넷 대회에서 힌튼 교수팀이 딥러닝 기술을 이용한 알렉스넷(Alexnet)으로 15.31%의 오류율을 기록하면서, 2등(26.2%)과 압도적인 차이로 우승함으로써 딥러닝 기술이 인공지능에 있어서 대세가 되었다. 딥 러닝의 경우, 입력된 데이터와 출력된 결과는 알 수 있지만 그 과정은 알 수 없는 경우가 많아 “Black Box Machine”이라 칭해지기도 한다. 인공지능 관련 기술이 발달함에 따라 현재 특정 영역에서 이용되고 있는 약한 인공지능(전문가 시스템)이 장차 통합되면서, 언젠가는 강한 인공지능(일반지능)이 출현될 것으로 예상된다. 약한 AI가 탑재된 로봇, 예를 들면 완전자율주행자동차에 법인격을 부여하기는 쉽지 않을 것 같다. 그러나 사회적 로봇에 대하여는 일정한 범위 내에서 동물과 유사한 지위를 부여할 수 있을 것 같다. 강한 AI가 탑재된 지능형 로봇의 경우 법인격을 부여받을 조건을 갖추었다고 볼 수 있다. 그러나 강한 AI의 출현이 인류의 재앙이 되지 않고 축복이 되기 위해서는, 관련 법률뿐만 윤리코드 등 연성법(soft law)에 대한 정비도 필요하다. Soon after Professor Hinton's team recorded an error rate of 15.31% with Alexnet using deep learning technology and won by an overwhelming difference from the second place who recorded an error rate of 26.2% in the ImageNet Challenge of 2012, deep learning technology became a trend in artificial intelligence. In the case of deep learning, input data and output results can be known, but the process is often not known, so it is called “Black Box Machine”. As artificial intelligence-related technologies develop, weak artificial intelligence (expert system) currently used in specific areas will be integrated in the future, and strong artificial intelligence (general intelligence) is expected to emerge one day. It will not be easy to give legal personality to robots equipped with weak AI, such as fully autonomous vehicles. However, it seems that social robots can be given a status similar to that of animals within a certain range. Intelligent robots equipped with strong AI can be said that they have met the conditions for to be granted legal personality. In order for the advent of the intelligent robots to become a blessing rather than a disaster for mankind, it is necessary to improve not only related laws but also soft laws such as ethical codes.

Exploring the Intersection of Artificial Intelligence in the Evolving Landscape of Art

한창희 한양대학교(ERICA캠퍼스) 한국미래문화연구소 2023 미래문화 Vol.0 No.8

This paper scrutinizes the collaborative role of AI as an artistic partner, examining how generative algorithms and neural networks contribute to the ideation phase of artistic creation. The evolution of generative art, driven by AI algorithms, emerges as a distinct genre within the artistic realm, offering novel and unpredictable expressions. A focal point of the investigation revolves around the key characteristics and features of AI arts, pondering the definition of AI arts and dissecting the innovations introduced in artistic creation through AI. The paper critically reflects on AI's impact on traditional artistic principles, prompting a reevaluation of concepts like authorship, originality, and the role of the artist. Looking toward the future, the inquiry concludes with reflections on the forthcoming landscape of art in the era of artificial intelligence. While acknowledging the unprecedented opportunities presented by AI in revolutionizing the art world, the paper highlights concerns related to biases, authenticity, and the potential displacement of human creators. The academic journey advocates for a measured and cautious approach, emphasizing the necessity of balancing the benefits of AI with the preservation of the unique qualities inherent in human-created art.