인공 지능에서 인공 감정으로‒감정을 가진 기계는 실현가능한가?‒

천현득 한국철학회 2017 철학 Vol.0 No.131

As artificial intelligence outperforms humans in some cognitive tasks which have been regarded as properly belonging to the human being, there is growing anxiety that people experience. Now, many people attempt to place the uniqueness of the humans in emotion rather than reason or intelligence. Since artificial emotions and emotional robots are emerging as new topics in artificial intelligence and robotics, however, philosophical treatment of them is required. I begin with discussing what has motivated to develop emotional robots and why emotions matter in robotics. After reviewing the current status of emotional robots, I examine whether the emotion-possessing robots are possible. The possibility of artificial emotions depends on what emotions really are. Instead of stipulating the definition of emotions, I suggest criteria on which we can assign emotions to other people, animal, or alien objects including artificial agents by considering the key roles played by emotions in intelligent beings. On these criteria, I argue, it is unlikely for genuinely emotional robots to be realized in the near future. Still, even if emotional robots would not appear shortly, unidirectional emotional bonds with social robots equipped with some degree of autonomy are potentially detrimental, which ought to be seriously considered in an ethical discussion of robots. 인공 감정에 관한 철학적 탐구가 필요한 시점이다. 인간의 고유한 영역으로 간주되던 인지적 과제에서 기계의 추월을 염려하는 처지가 되자, 사람들은 이제 이성이 아니라 감정에서 인간의 고유성을 찾으려한다. 하지만 최근인공지능 로봇에 감성을 불어넣는 작업이 새로운 화두로 떠오르고 있다. 이 글은 인공 감정의 실현가능성과 잠재적인 위험을 논의한다. 먼저, 감성 로봇의 개발 현황과 주요한 동기들을 개괄하고, 왜 로봇의 감정이 문제가 되는지 살펴본다. 진정한 감정-소유 로봇이 가능한지 검토하기 위해 감정을 선험적으로 정의하기보다는 감정이 수행하는 몇 가지 핵심 역할을 소개하고 이로부터 어떤 대상에 감정을 부여할 수 있는 기준들을 제안한다. 나는 이런 기준에 비추어 진정한 감정 로봇이 근미래에 실현될 가능성이 낮다고 주장한다. 그러나 감정-소유로봇이 등장하기 이전이라도, 어느 정도의 자율성을 가진 로봇과 맺는 일방적감정 소통은 잠재적으로 위험할 수 있으며, 이에 대비하는 것이 시급하다고 주장한다.

인공정서를 가진 스마트폰용 인터랙티브 캐릭터 개발

함준석 ( Jun Seok Ham ),여지혜 ( Ji Hye Yeo ),박성호 ( Sung Ho Park ),고일주 ( Il Ju Ko ) 한국감성과학회 2011 감성과학 Vol.14 No.4

This paper purposes to develop an artificial emotion reflecting emotional features contains situations, time, and characteristics, also to develop an interactive character having this artificial emotion with a smart phone. The artificial emotion has an Emotion Module and Drive Module for expressing emotion according to external emotional stimulus and internal drive. The Emotion Module administrates emotions according to time, characteristic, interrelation between different emotions. The Drive Module controls sensitivities of emotion according to changing drives over long time. Also due to defence mechanism for expressing emotions, emotions are processed by two pathways: The first pathway which is affected by the Emotion Module and the Drive Module, and the second pathway that is not to be done. We developed an interactive character having the artificial emotion with this structure using smart phone. And we simulated the artificial emotion what differences there are according to situations, characteristic, and time under same input conditions. The result of this paper has meanings developing the interactive character having the artificial emotion actually, and making it possible to personalize an artificial emotion with expressing the artificial emotion using smart phone.

초등 인공지능 교육을 위한 핵심역량 모델링

조수경,최미설 핵심역량교육학회 2022 핵심역량교육연구 Vol.7 No.1

The purpose of this study is to clarify the concept of talent for the artificial intelligence era and to provide the foundation for designing the directivity of artificial intelligence education by modeling core competencies for upper grades of elementary school students where artificial intelligence education is initiated in public education. To this end, experts one-to-one interviews, focus group interviews, and Delphi surveys were repeatedly conducted, in which process, core competency components of elementary artificial intelligence education were derived and validated. The core competencies of artificial intelligence education included 15 competencies in the ‘artificial intelligence understanding’ category, 8 competencies in the ‘artificial intelligence ethics’ category, and 8 competencies in the ‘artificial intelligence social emotions’ category, resulting in a total of 31 competencies in 3 categories. As a result of the Delphi survey on the final draft, the overall response average was 3.66/4 points, the CVI average was 0.930, and the CVR average was 0.863, which exceeded the CVI minimum value of 0.70 and the CVR minimum value of 0.455. In particular, the validity and importance of the competencies of 'AI awareness', 'AI utilization' in the 'artificial intelligence understanding' category, ‘AI ethics problem recognition’ in the 'artificial intelligence ethics’ category, and 'human-AI collaboration', 'active self-management' in the 'artificial intelligence social emotions' category were found to be high. This study attempted to contribute to expanding the perspective beyond artificial intelligence principle education to artificial intelligence ethics and social-emotional education through modeling of core competencies for elementary schoolers and presented the basis for the systematic perspective of artificial intelligence education. It is also meaningful to lay the foundation for introducing competency-based convergence education into artificial intelligence education. 본 연구는 인공지능 교육이 공교육에 도입되기 시작하는 초등학교 고학년을 대상으로 인공지능 핵심역량을 모델링함으로써 인공지능 시대의 인재상을 명확히 하고, 인공지능 교육의 방향성 수립에 기초를 제공하는 것을 목적으로 한다. 이를 위하여 전문가 일대일 면담, 포커스 그룹 인터뷰, 델파이 조사 등 의견수렴 및 타당화 과정을 반복적으로 실시하였고, 이러한 과정에서 초등 인공지능 교육의 핵심역량 구성요인을 도출하고 타당성을 검증하였다. 인공지능 교육 핵심역량은 ‘인공지능 이해’ 범주에 15개 역량, ‘인공지능 윤리’ 범주에 8개 역량, ‘인공지능 사회정서’ 범주에 8개 역량이 포함되어 총 3개 범주 31개 역량이 도출되었다. 최종안에 대한 델파이 조사결과, 전체 응답평균 3.66/4점, CVI 평균 0.930, CVR 평균 0.863으로 CVI 최솟값 0.70과, CVR 최솟값 0.455를 상회하는 등 높은 공감대를 획득하였다. 특히, ‘인공지능 이해’ 범주의 ‘AI 인지’와 ‘AI 활용’, ‘인공지능 윤리’ 범주의 ‘AI 윤리 문제인식’, ‘인공지능 사회정서’ 범주의 ‘인간-AI 협업’과 ‘능동적 자기관리’ 역량의 타당도와 중요도가 높게 나타났다. 본 연구는 초등 맥락의 인공지능 교육 핵심역량 모델링을 통해 인공지능 이해 교육을 넘어 인공지능 윤리와 사회정서 교육까지로 시야를 넓히는데 기여하고자 하였으며, 초중등 인공지능 교육의 체계를 마련하는데 기초를 제시하였다. 또한 인공지능 교육에 역량중심 융합교육을 도입할 수 있도록 토대를 마련한 것에 의의를 가진다.

인공 지능에서 인공 감정으로

천현득(Hyundeuk Cheon) 한국철학회 2017 철학 Vol.0 No.131

인공 감정에 관한 철학적 탐구가 필요한 시점이다. 인간의 고유한 영역으로 간주되던 인지적 과제에서 기계의 추월을 염려하는 처지가 되자, 사람들은 이제 이성이 아니라 감정에서 인간의 고유성을 찾으려한다. 하지만 최근 인공지능 로봇에 감성을 불어넣는 작업이 새로운 화두로 떠오르고 있다. 이 글은 인공 감정의 실현가능성과 잠재적인 위험을 논의한다. 먼저, 감성 로봇의 개발 현황과 주요한 동기들을 개괄하고, 왜 로봇의 감정이 문제가 되는지 살펴본다. 진정한 감정-소유 로봇이 가능한지 검토하기 위해 감정을 선험적으로 정의하기보다는 감정이 수행하는 몇 가지 핵심 역할을 소개하고 이로부터 어떤 대상에 감정을 부여할 수 있는 기준들을 제안한다. 나는 이런 기준에 비추어 진정한 감정 로봇이 근미래에 실현될 가능성이 낮다고 주장한다. 그러나 감정-소유 로봇이 등장하기 이전이라도, 어느 정도의 자율성을 가진 로봇과 맺는 일방적 감정 소통은 잠재적으로 위험할 수 있으며, 이에 대비하는 것이 시급하다고 주장한다. As artificial intelligence outperforms humans in some cognitive tasks which have been regarded as properly belonging to the human being, there is growing anxiety that people experience. Now, many people attempt to place the uniqueness of the humans in emotion rather than reason or intelligence. Since artificial emotions and emotional robots are emerging as new topics in artificial intelligence and robotics, however, philosophical treatment of them is required. I begin with discussing what has motivated to develop emotional robots and why emotions matter in robotics. After reviewing the current status of emotional robots, I examine whether the emotion-possessing robots are possible. The possibility of artificial emotions depends on what emotions really are. Instead of stipulating the definition of emotions, I suggest criteria on which we can assign emotions to other people, animal, or alien objects including artificial agents by considering the key roles played by emotions in intelligent beings. On these criteria, I argue, it is unlikely for genuinely emotional robots to be realized in the near future. Still, even if emotional robots would not appear shortly, unidirectional emotional bonds with social robots equipped with some degree of autonomy are potentially detrimental, which ought to be seriously considered in an ethical discussion of robots.

인공지능과 감정지능 - 누스바움의 감정론을 중심으로 -

이선 ( Lee Sun ) 전북대학교 인문학연구소 2017 건지인문학 Vol.0 No.19

This paper tries to elucidiate what is emotional intelligence in the age of Artificial Intelligence. First of all, this paper considers possibilities and limits of Artificial Intelligence. Artificial Intelligence can`t learn the ability to feel. According to Martha C. Nussbaum`s theory of emotions, emotions are cognitive and valuable judgment. In the one hand, emotions are cognitiv, because emotions perceive external goods in despites of uncontrollability. In the other hand, emotions are valuable, because the judgment of emotions ist oriented to external objects and the world. Furthermore, she defines emotions as eudaimonistic because they are concerned with the flourishing and well-being of the agent. The eudaimonistic judgment of emotions constitutes an ethical form of emotions compared to Aristotle`s eudaimonia. Finally, this paper seeks to show a new way of intelligence through emotional in-telligence in Nussbaum`s theory of emotions.

인공지능 이슈에 대한 뉴스 프레임이 이용자의 정서와 행동 의향에 미치는 영향

장미경(Mikyung Chang),민영(Young Min) 한국언론학회 2024 한국언론학보 Vol.68 No.4

The purpose of this study is to identify social and policy implications by exploring the characteristics and effects of news frames related to artificial intelligence (AI) technologies, which have emerged as a global issue. To achieve this goal, 〈Study 1〉 conducted manual and computer-assisted content analyses of the AI-related news frames of major domestic news outlets in terms of value frames (progress versus crisis) and attributional frames (high versus low controllability). In 〈Study 2〉, an online survey experiment was conducted using a 2×2 factorial design to explore the effects of news framing on users’ emotions and behavioral intentions. The key findings are as follows. First, the news media tended to use the ‘progress value’ frames more frequently than the ‘crisis value’ frames, focusing on the social developments and benefits AI technologies could bring. In terms of controllability, the news media more often emphasized sufficient levels of human control over the technologies. Second, the news frames related to AI technologies elicited differentiated emotional responses; progress value frames evoked positive emotions such as hope and pride, whereas crisis value frames strengthened negative emotions such as anger. The controllability frames did not significantly moderate the effects of value frames on emotions. Third, the emotions elicited by news frames were further associated with various behavioral intentions. Specifically, the emotions of hope and pride enhanced the intention to engage in information-seeking activities to acquire AI-related knowledge and information. The feeling of anger, on the contrary, significantly strengthened users’ intention for critical activities to express their opinion on the side effects of AI technologies, urging the attention of the media, government, and academia. Additionally, when users felt anger, their intention to engage in information-seeking behavior decreases while their intention to avoid or reject the use of AI products and services increases. Fourth, it was consequently confirmed that emotions significantly mediated the relationship between news frames and behavioral intentions, demonstrating that how individuals feel about AI news affects their subsequent actions. The combination of content analysis and experimental design enhanced the study’s validity and reliability, providing a comprehensive understanding of how news framing influences the public’s emotional responses and behavior regarding AI. The study underscores the importance of emotions in the acceptance of new technologies and validates the use of framing theory in science communication research to explore and understand public discourse and reactions to critical science issues. This study also highlights the need for balanced media coverage and the creation of public forums for discussion and policy development concerning new technologies. Overall, the study calls for greater attention to emotional responses in technology acceptance and the need for responsible media practices to foster informed public discourse on AI.

심리학 기반 인공감정모델을 이용한 감정의 시각화 -햄릿의 등장인물을 중심으로-

함준석 ( Jun Seok Ham ),여지혜 ( Ji Hye Ryeo ),고일주 ( Il Ju Ko ) 한국감성과학회 2008 감성과학 Vol.11 No.4

We cannot express emotions correctly with only speech because it is hard to estimate the kind, size, amount of emotions. Hamlet who is a protagonist in `Hamlet` of Shakespeare has emotions which cannot be expressed within only speech because he is in various dramatic situations. So we supposed an artificial emotion, instead of expressing emotion with speech, expressing and visualizing current emotions with color and location. And we visualized emotions of characters in `Hamlet` with the artificial emotion. We designed the artificial emotion to four steps considering peculiarities of emotion. First, the artificial emotion analyzes inputted emotional stimulus as relationship between causes and effects and analyzes its kinds and amounts. Second, we suppose Emotion Graph Unit to express generating, maintaining, decaying of analyzed one emotional stimuli which is outputted by first step, according to characteristic. Third, using Emotion Graph Unit, we suppose Emotion Graph that expresses continual same emotional stimulus. And we make Emotion Graph at each emotions, managing generation and decay of emotion individually. Last, we suppose Emotion Field can express current combined value of Emotion Graph according to co-relation of various emotions, and visualize current emotion by a color and a location in Emotion Field. We adjusted the artificial emotion to the play `Hamlet` to test and visualize changes of emotion of Hamlet and his mother, Gertrude.

AI 기술을 이용한 관객 상호작용 노화현상 생성 미디어아트 - 연구자의 미디어아트 작품 〈Chemistry〉를 중심으로 -

마적 ( Ma¸ Di ),김형기 ( Kim¸ Hyunggi ) 한국영상미디어협회 2020 예술과 미디어 Vol.19 No.1

본 논문은 상호주관성의 철학적 사유에 입각하여 인간과 인공지능의 의식과 감정 사이의 관계를 탐구한다. 제2장에서는 국내외 미디어아트와 생성적 대립 신경망(Generative Adversarial Network)의 기술 개발 현황, 그리고 작품 〈Chemistry〉의 예술성과 기술성에 대해 자세히 설명했다. 제3장에서는 작품 〈Chemistry〉가 주는 영감과 사고를 바탕으로 한 상호주관성 사유의 관점에서 인공지능이 ‘인간 의식과 감정의 확장’이라는 점을 철학적으로 증명하고자 시도했다. 인공지능이란, 인간의 지능을 기술적으로 모방하여 ‘사물화’시킨 것이다. 본 작품에서는 이러한 인공지능을 활용하여 감상하는 관객이 공감하도록 유도했다. 대략적으로 설명하자면, 우선 인공지능으로 사람의 얼굴을 인식하여 안면 노화를 진행시킨다. 그리고 사계절의 변화와 꽃이 시드는 과정을 함께 보여줌으로써 관객은 마음 깊은 곳에서 작품과 상호작용 할 수 있도록 했다. 이러한 과정을 통해 ‘인간의 의식’에서는 확장이 일어난다. 제3장에서는 인공지능의 단점을 보완하기 위한 해법을 제시했다. 왜냐하면 오늘날 아무리 빠르게 발전하고 있는 인공지능이라고 해도 인간의 깊은 감정적 욕구까지 충족하지 못하기 때문이다. 이때 예술이 인간과 기계 사이에서 일어나는 상호작용의 매개체로 매우 중요한 역할을 한다. 인공지능은 인간의 심층적인 감정과 그 울림을 정확하게 식별하지 못한다. 이에 반해 예술은 가장 인간적인 요소이다. 인간이 창조한 예술은 무형의 형태로 모든 사람들에게 지속적인 영향을 미치고 있다. 예술은 매우 간단해 보이는 방식으로도 사람의 마음을 감동시킬 수 있다. 필자의 작품에서는 예술을 인간과 인공지능 간의 소통을 가능케 하는 ‘다리’ 역할을 한다. 과학 연구에 대한 사회적 관심이 날로 높아지는 가운데, 우리는 철학적 관점에서 인간과 인공지능 사이의 관계를 깊이 고찰해야 하며 매개체 역할을 하는 인공지능의 능동성에 주목해야 한다. This article uses the philosophy of intersubjectivity to examine the connection between artificial intelligence and human emotional awareness. It first analyzes Korea and overseas new media art along with technological development based on generative adversarial networks, and then provides a detailed explanation of the artistry and techniques used in the new media artwork Chemistry. while the third chapter, on the foundation of our grouped artwork as inspirational thinking and based on intersubjective thinking, elaborates on artificial intelligence as a philosophical viewpoint of the extension of human emotional awareness. Artificial intelligence is a simulated technological phase of awareness and the ‘objectification’ of intelligence. People create a sense of empathy when they view artwork. In addition, artificial intelligence can distinguish between human faces, and can detect signs of aging, seasonal changes, as well as when flowers wilt. Artwork causes people to react with profound emotional responses that expand their human awareness. The third chapter finally proposes, in today’s era of rapidly developing artificial intelligence, how artificial intelligence can fulfill humanity’s defects displayed through their deep emotions. The following chapter proposes a solution for this problem, which is to turn art into an emotion- based medium that can enable communication between human beings and artificial intelligence. Artificial intelligence is incapable of precisely and effectively differentiating between humans’ deep emotional fluctuations. In contrast, art is one of the most humanized things. Humanity creates art, but at the same time, art’s intangible thinking constantly influences each person. It can even use what seems like the simplest methods to touch people’s soul. Therefore, our artwork will view art as an ‘intermediary’ to enable people to communicate with artificial intelligence, and in today’s era in which all of society focuses more on scientific research, this article tries to emphasize the importance of art’s development in artificial intelligence.

드라마 대본 분석을 통한 등장인물의 성격이 반영된 인공정서 캐릭터 개발 - ``솔약국집 아들들``을 중심으로

함준석 ( Jun Seok Ham ),이신영 ( Shin Young Rhee ),방그린 ( G Reen Bang ),고일주 ( Ii Ju Ko ) 한국감성과학회 2012 감성과학 Vol.15 No.2

This paper looks to extract personality traits from the drama characters within a drama script, and to apply it them to a character that has an artificial emotion. The method of applying the personality of a character from a drama script is as follows. First, we separate a drama script into several pieces, by the characters therin. Next, we extract emotion-related terms by matching morphemes analysis and by using an emotion terms database. Next, we analyze a dominant emotion using extracted emotion terms. Finally last, we apply the analyzed dominant emotion to an equation pertaining to artificial emotion. We made progress in developing user evaluation that features blind testing, to verify that the artificial emotion character bears the personality of a drama character. We apply three drama character personalities to artificial emotion characters bearing the same appearance. The user had to match three artificial emotion characters and drama characters according to personality. The users had a high percentage of correct answers, thus confirming the efficacy of our method of applying a personality, using information from a drama script.

인공지능과 감정지능 – 누스바움의 감정론을 중심으로 –

이선 전북대학교 인문학연구소 2017 건지인문학 Vol.0 No.19

This paper tries to elucidiate what is emotional intelligence in the age of Artificial Intelligence. First of all, this paper considers possibilities and limits of Artificial Intelligence. Artificial Intelligence can’t learn the ability to feel. According to Martha C. Nussbaum’s theory of emotions, emotions are cognitive and valuable judgment. In the one hand, emotions are cognitiv, because emotions perceive external goods in despites of uncontrollability. In the other hand, emotions are valuable, because the judgment of emotions ist oriented to external objects and the world. Furthermore, she defines emotions as eudaimonistic because they are concerned with the flourishing and well-being of the agent. The eudaimonistic judgment of emotions constitutes an ethical form of emotions compared to Aristotle’s eudaimonia. Finally, this paper seeks to show a new way of intelligence through emotional intelligence in Nussbaum’s theory of emotions.