DOI QR코드

DOI QR Code

What Concerns Does ChatGPT Raise for Us?: An Analysis Centered on CTM (Correlated Topic Modeling) of YouTube Video News Comments

ChatGPT는 우리에게 어떤 우려를 초래하는가?: 유튜브 영상 뉴스 댓글의 CTM(Correlated Topic Modeling) 분석을 중심으로

  • Song, Minho (Dept. of Media & Communication, Incheon National University ) ;
  • Lee, Soobum (Dept. of Media & Communication, Incheon National University)
  • Received : 2024.02.02
  • Accepted : 2024.03.08
  • Published : 2024.03.31

Abstract

This study aimed to examine public concerns in South Korea considering the country's unique context, triggered by the advent of generative artificial intelligence such as ChatGPT. To achieve this, comments from 102 YouTube video news related to ethical issues were collected using a Python scraper, and morphological analysis and preprocessing were carried out using Textom on 15,735 comments. These comments were then analyzed using a Correlated Topic Model (CTM). The analysis identified six primary topics within the comments: "Legal and Ethical Considerations"; "Intellectual Property and Technology"; "Technological Advancement and the Future of Humanity"; "Potential of AI in Information Processing"; "Emotional Intelligence and Ethical Regulations in AI"; and "Human Imitation."Structuring these topics based on a correlation coefficient value of over 10% revealed 3 main categories: "Legal and Ethical Considerations"; "Issues Related to Data Generation by ChatGPT (Intellectual Property and Technology, Potential of AI in Information Processing, and Human Imitation)"; and "Fear for the Future of Humanity (Technological Advancement and the Future of Humanity, Emotional Intelligence, and Ethical Regulations in AI)."The study confirmed the coexistence of various concerns along with the growing interest in generative AI like ChatGPT, including worries specific to the historical and social context of South Korea. These findings suggest the need for national-level efforts to ensure data fairness.

본 연구는 ChatGPT로부터 촉박된 생성형 인공지능에 대해 국내의 특수성을 고려한 대중의 우려를 살펴보고자 하였다. 이를 위해 유튜브에서 102개의 윤리 관련 뉴스 영상에 포함된 댓글을 파이썬 스크래퍼를 개발하여 수집하였으며, 텍스톰을 통해 형태소 분석 및 전처리를 통해 15,735개 댓글을 대상으로 상관토픽모델(CTM)을 통해 분석하였다. 분석 결과, 뉴스 영상에 포함된 댓글의 주요 토픽은 '법적 및 윤리적 고려 사항', '지적 재산권 및 기술', '기술 발전과 인류 미래, 정보 처리에서 인공지능의 잠재력', 'AI에서의 감정 지능 및 윤리적 규제', '인간모방' 등 6개로 확인되었다. 또한 6개의 토픽을 10% 이상의 상관계수 값을 보이는 관계로 구조화한 결과 '법적 및 윤리적 고려 사항', 'ChatGPT의 데이터 생성 관련 이슈(지적 재산권 및 기술, 정보 처리에서의 인공지능의 잠재력, 인간모방', '인류 미래에 대한 두려움(기술 발전과 인류 미래, AI에서의 감정 지능 및 윤리적 규제)' 등 3개로 구조화할 수 있었다. 이를 바탕으로 ChatGPT로 인해 촉발된 생성형 인공지능에 관한 관심과 더불어 다양한 우려가 공존하고 있는 것을 확인하였고, 국내의 역사적 및 사회적 맥락을 반영한 특수성을 가진 우려도 존재하고 있음을 확인하였다. 이러한 결과를 통해 데이터 공정성에 대한 국가 주도의 노력이 필요함을 제안하였다.

Keywords

Acknowledgement

This work was supported by Incheon National University Research Grant in 2022

References

  1. Alasadi, E. A. & Baiz, C. R. (2023). "Generative AI in Education and Research: Opportunities, Concerns, and Solutions." Journal of Chemical Education, 100(8), 2965-2971.
  2. Alawida, M., Mejri, S., Mehmood, A., Chikhaoui, B. & Isaac Abiodun, O. (2023). "A Comprehensive Study of ChatGPT: Advancements, Limitations, and Ethical Considerations in Natural Language Processing and Cybersecurity." Information, 14(8), 462.
  3. Al-Khalifa, S., Alhumaidhi, F., Alotaibi, H., & Al-Khalifa, H. S. (2023). "ChatGPT across Arabic Twitter: A Study of Topics, Sentiments, and Sarcasm." Data, 8(11), 171.
  4. Ann, M. (2017). "A Qualitative Research on Internet Discourse on Artificial Intelligence (AI): Based on YouTube Video Comments." The Society of Convergence Knowledge Transactions, 5(2), 63-69. 
  5. Arabshahi, F. & Anandkumar, A. (2017). "Spectral Methods for Correlated Topic Models." Paper presented at The International Conference on Artificial Intelligence and Statistics, 20-22 April.
  6. Bahroun, Z., Anane, C., Ahmed, V. & Zacca, A. (2023). "Transforming Education: A Comprehensive Review of Generative Artificial Intelligence in Educational Settings through Bibliometric and Content Analysis." Sustainability, 15(17), 12983.
  7. Bendel, O. (2023). "Image SynThesis from an Ethical Perspective." AI & SOCIETY, 1-10.
  8. Bickley, S. J. & Torgler, B. (2023). "Cognitive Architectures for Artificial Intelligence Ethics." AI & SOCIETY, 38(2), 501-519.
  9. Blei, D. M. & Lafferty, J. D. (2007). "A Correlated Topic Model of Science." Paper presented at the proceedings of The Annals of Applied Statistics, 1 June.
  10. Blei, D. & Lafferty, J. (2006). "Correlated Topic Models." Advances in Neural Information Processing Systems, 18, 147-154.
  11. Bukar, U., Sayeed, M. S., Razak, S. F. A., Yogarayan, S. & Amodu, O. A. (2023). "Text Analysis of Chatgpt as a Tool for Academic Progress or Exploitation." Available at SSRN 4381394.
  12. Cascella, M., Montomoli, J., Bellini, V. & Bignami, E. (2023). "Evaluating The Feasibility of ChatGPT in Healthcare: An Analysis of Multiple Clinical and Research Scenarios." Journal of Medical Systems, 47(1), 1-5.
  13. Chang, J., Boyd-Graber, J., Gerrish, S., Wang, C. & Blei, D. M. (2009). "Reading Tea Leaves: How Humans Interpret Topic Models." Advances in Neural Information Processing Systems 22, 288-296.
  14. Chatterjee, J. & Dethlefs, N. (2023). "This New Conversational AI Model Can be Your Friend, Philosopher, and Guide... and even Your Worst Enemy." Patterns, 4(1). 1-3.
  15. Choi, H., Lee, J. & Chin, M. (2020). "Intimate Partnerships and Family Policy in Korean News Articles and Comments: A Topic Model Analysis." Family and Culture, 32(4), 29-60.  https://doi.org/10.21478/FAMILY.32.4.202012.002
  16. Choi, J. (2023). "A Study on Artificial Intelligence and Disinformation." Media and Personality Rights, 9(2), 37-82. 
  17. Choi, J., Jung, Y. & Kim, J. (2023). "Analysis of Youtube News Contents on College Performance Evaluation and Consulting: Using Mixed Methodology of Content Analysis and Topic Modeling." Locality & Communication, 27(4), 123-169. 
  18. Choi, J. & Schwarcz, D. (2023). "AI Assistance in Legal Analysis: An Empirical Study." Available at SSRN 4539836.
  19. Choi, J. & Lee, H. (2023). Digital News Report 2023 Korea. Seoul: Korea Press Foundation. 
  20. Chu, H., Cho, W. & Yoo, J. (2023). "A Study on The Methodology for Exploring AI Trustworthiness Policy Related Issues using Language Models and Text Mining." Journal of Korea Technology Innovation Society, 26(4), 572-593. 
  21. Chung, N. (2023). "Recent Trends and Implications of Regulating Artificial Intelligence in The Digital Transformation of Public Administration." Media and Personality Rights, 9(2), 1-36. 
  22. Ciston, S. (2019). "Intersectional AI is Essential: Polyvocal, Multimodal, Experimental Methods to Save Artificial Intelligence." Journal of Science and Technology of The Arts, 11(2), 3-8.
  23. Eke, D. O. (2023). "ChatGPT and The Rise of Generative AI: Threat to Academic Integrity?." Journal of Responsible Technology, 13, 100060.
  24. Eom, K. & Kim. D. (2021). "Is Political Polarization Reinforced in the Online World?: Empirical Findings of Comments about News Articles." Informatization Policy, 28(4), 19-35.  https://doi.org/10.22693/NIAIP.2021.28.4.019
  25. Farrelly, T. & Baker, N. (2023). "Generative Artificial Intelligence: Implications and Considerations for Higher Education Practice." Education Sciences, 13(11), 1109.
  26. Ferrara, E. (2023). "GenAI Against Humanity: Nefarious Applications of Generative Artificial Intelligence and Large Language Models." Journal of Computational Social Science, 6(2), 1-21. https://doi.org/10.2139/ssrn.4614223
  27. Fui-Hoon Nah, F., Zheng, R., Cai, J., Siau, K. & Chen, L. (2023). "Generative AI and ChatGPT: Applications, Challenges, and AI-human Collaboration." Journal of Information Technology Case and Application Research, 25(3), 277-304.
  28. Gilson, A., Safranek, C. W., Huang, T., Socrates, V., Chi, L., Taylor, R. A. & Chartash, D. (2023). "How Does ChatGPT Perform on The United States Medical Licensing Examination? The Implications of Large Language Models for Medical Education and Knowledge Assessment." JMIR Medical Education, 9(1), e45312.
  29. Guo, J. & Jeon, J. (2023). "The Effect of Comprehensive Thinking Tendency, Perceived Usefulness, and Artificial Intelligence Ethical Risk Perception on Attitude and Intention to Use Artificial Intelligence Technology." Culture & Convergence, 45(5), 521-536. 
  30. Heo, E., Lee, Y. & Shim, J. (2020). "Why Ethics is: A Landscape of Modern AI Ethics Debate, Its Features and Limitations." Human Beings, Environment and Their Future, 24, 165-209. 
  31. Hutson, M. (2021). "Robo-Writers: The Rise and Risks of Language-Generating AI." Nature, 591(7848), 22-25. https://doi.org/10.1038/d41586-021-00530-0
  32. Jang, Y. (2023). "An Examination of The Function and Impact of User Comments as Cues to Public Opinion." Communication Theories, 19(3), 5-53.  https://doi.org/10.20879/ct.2023.19.3.005
  33. Jobin, A., Ienca, M. & Vayena, E. (2019). "The Global Landscape of AI Ethics Guidelines." Nature machine intelligence, 1(9), 389-399.
  34. Jung, S. (2022). "European Parliament Resolution on The Use of Artificial Intelligence in Criminal Justice: Regarding The Ban of Automated Decisions by Artificial Intelligence." Korean Journal of Criminology, 34(2), 137-172. 
  35. Kangaspunta, V. (2016). "Talvivaara ja Kommenteissa Keskustelevat Julkisot: Keskustelunanalyyttinen Menetel makokeilu Verkkouutisten Kommentoinnista." Media & viestinta, 39(1), 24-54.
  36. Kangaspunta, V. (2018). "Online News Comments: Social Network and Emergent Public." The Information Society, 34(5), 275-288. https://doi.org/10.1080/01972243.2018.1497741
  37. Kim, D. & Hong, H. (2015). "The Limited Meanings and Roles of The Reply as The Instrument of Political Participation." Journal of Cybercommunication Academic Society, 32(1), 51-86. 
  38. Kim, H. (2019). "Exploring Information Ethics Issues based on Text Mining using Big Data from Web of Science." The Journal of Korean Association of Computer Education, 22(3), 67-78. 
  39. Kim, J., Kang, S. & Kwon, S. (2023). "A Study on The Recognition of Teacher Librarians on The Introduction of ChatGPT in School Library." Journal of The Korean Society for Library and Information Science, 57(2), 349-377. 
  40. Kim, M. & Park, S. (2023). "Analysis of AI Ethics Research Trends using Text Mining." Journal of The Korean Association of Information Education, 27(1), 23-33. 
  41. Kim, M., Park, S. & Roh, S. (2022). "A Study on The Awareness of Artificial Intelligence Development Ethics based on Social Big Data." Engineering Education Research, 25(3), 35-44.
  42. Kim, S. (2021). "The Latest Trends of AI Legislation and The Requirements for its Establishment: Focused on Comparison with AI Strategies and Legislative Acts of EU." Public Law Journal, 22(4), 115-148.  https://doi.org/10.31779/plj.22.4.202111.005
  43. Kim, T. & Han, S. (2023). "An Exploratory Study on Social Issues Related to ChatGPT: Focusing on News Big Data-based Topic Modeling Analysis." Journal of Digital Contents Society, 24(6), 1209-1220. 
  44. Ko, M., Jun, C. & Song, H. (2023). "A Study on The Korean Public's Perception of ChatGPT through Text Mining Analysis." Journal of The Korea Contents Association, 23(4), 1-13. 
  45. Ksiazek, T. B., Peer, L. & Lessard, K. (2016). "User Engagement with Online News: Conceptualizing Interactivity and Exploring The Relationship between Online News Videos and User Comments." New media & society, 18(3), 502-520.
  46. Kwon, J. & Lee, Y. (2023). "Critical Review of Government Policy on AI Education in The ChatGPT Era." Artificial Intelligence Humanities Research, 13, 9-38. 
  47. Kwon, K. & Kim, S. (2023). "Analysis of Socio-Economic Issues of ChatGPT based on LDA Topic Modeling." Journal of Korea Technology Innovation Society, 26(4), 594-613. 
  48. Lee, B. & Jho, H. (2023). "Analysis of Energy-related Research Topics at The National Science Fair using Topic Modeling and Conversational Artificial Intelligence (ChatGPT)." Journal of Energy and Climate Change Education, 13(2), 151-161. 
  49. Lee, E. (2011). "Perceived Bias or Biased Perception? Effects of OThers' Comments, Perceived Opinion Climate, and Issue Involvement on Perceived News Slant." Korean Journal of Journalism & Communication Studies, 55(3), 179-198. 
  50. Lee, H. (2023). "Ethical Guidelines for Controlling Bias of Artificial Intelligence Chatbots." Public Law, 51(3), 715-744.  https://doi.org/10.38176/PublicLaw.2023.2.51.3.715
  51. Lee, H. (2023). "The Necessity of Enacting AI Legislation and The Challenges in Administrative Law." Kookmin Law Review, 35(3), 9-57.  https://doi.org/10.17251/LEGAL.2023.35.3.9
  52. Lee, H. & Cheon, H. (2021). "The Principle of Explicability in AI Ethics." Study of Humanities, 35, 37-63. 
  53. Lee, H. & Ha, J. (2022). "Trends of Fairness and Transparency in Hyperscale Language Models." Communications of The Korean Institute of Information Scientists and Engineers, 40(11), 36-43. 
  54. Lee, K. (2021). "The Topology of AI-automated Platform Labor in The Phase of COVID-19." Korean Journal of Communication & Information, 109, 67-98. 
  55. Lee, S. (2016). "Is It Possible to be a Moral Artificial Intelligence?: The Problem of Moral and Legal Responsibility in A.I." Law and Policy Studies, 16(4), 293-303. 
  56. Lee, S., Song, J. & Kim, Y. (2010). "An Empirical Comparison of Four Text Mining Methods." Journal of Computer Information Systems, 51(1), 1-10.
  57. Lee, S. R. & Choi, E. (2022). "A Data Analysis and Visualization of AI Ethics: Focusing on The Interactive AI Service 'LeeLuda'." Journal of Digital Convergence, 20(2), 269-275. 
  58. Lee, Y. (2023). "AI Ethics for Large Language Models(LLMs): Based on ChatGPT." Paper presented at the proceedings Korean Institute of Information Scientists and Engineers, June, 17-20. 
  59. Mannuru, N. R., Shahriar, S., Teel, Z. A., Wang, T., Lund, B. D., Tijani, S., Pohboon, C. O., Agbaji, D., Alhassan, J., Galley, J., Kousari, R., Ogbadu-Oladapo, L., Saurav, S. K., Srivastava, A., Tummuru, S. P., Uppala, S. & Vaidya, P. (2023). "Artificial Intelligence in Developing Countries: The Impact of Generative Artificial Intelligence (AI) Technologies for Development." Information Development, 02666669231200628.
  60. McCambridge, L. (2022). "Describing The Voice of Online Bullying: An Analysis of Stance and Voice Type in YouTube Comments." Discourse, Context & Media, 45, 520-531.
  61. MacKenzie, D. & Wajcman, J. (1999). The social shaping of technology. Milton Keynes: Open University Press.
  62. Muthusami, R. & Saritha, K. (2022). "A Novelty Analysis about an Impact of Tweets and Twitter Bios on Topic Quality Discovery using The Topic Modeling." Journal of The Institution of Engineers (India): Series B, 103(5), 1431-1441.
  63. Nam, K. (2023). "A Study on Appropriateness of Risk-based Approach for The Purpose of Regulating AI." Kookmin Law Review, 36(1), 159-192. 
  64. Newman, D., Lau, J. H., Grieser, K. & Baldwin, T. (2010). Automatic evaluation of topic coherence. Paper presented at the proceedings The Human Language Technologies: The 2010 Annual Conference of The North American Chapter of The Association for Computational Linguistics, 2-4 June.
  65. Ng, C. W. (2018). "Critical Multimodal Discourse Analyses of News Discourse on Facebook and YouTube." Journal of Asia TEFL, 15(4), 1174-1185. https://doi.org/10.18823/asiatefl.2018.15.4.22.1174
  66. Noh, D. & Hong, M. (2023). "Strategies for Solving The AI Plagiarism Problem and Educational Applications of ChatGPT." The Journal of Korean Language and Literature Education, 82, 71-102. 
  67. Noh, M. (2022). "Analysis of Trends in 'AI Ethics' Research using Text Mining Techniques: Before and After The Announcement of National Artificial Intelligence Ethics Standards (draft), Focusing on Domestic Journals." Journal of AI Ethics, 1(1), 72-97.  https://doi.org/10.59728/JAIE.2022.1.1.72
  68. O'callaghan, D., Greene, D., Carthy, J. & Cunningham, P. (2015). "An Analysis of The Coherence of Descriptors in Topic Modeling." Expert Systems with Applications, 42(13), 5645-5657.
  69. Oh, H. & Cheong. Y. (2021). "A Study on The Charateristics of eWOM of Consumers as YouTube Creators Using Correlated Topic Model." The Korean Journal of Advertising and Public Relations, 23(3), 37-72. 
  70. Oh, S. (2023). A Study on The Case of Using ChatGPT & Learners' Perceptions in College Liberal Arts Writing. Korean Journal of General Education, 17(3), 11-23. 
  71. Oh, S., Jang, M. & Park, J. (2023). "Undergraduates' Awareness of The Ethics of Generative AI Utilization in College Writing." The Korean Journal of Literacy Research, 14(4), 69-96. 
  72. Park, K. (2022). "AI Bias and Citizenship." Social Studies Education, 61(2), 95-106.  https://doi.org/10.37561/sse.2022.06.61.2.95
  73. Park, S. & Lee, J. (2023). "Artificial Intelligence Fact-Checking Technology and The Dynamics of Factuality: An In-depth Interview Analysis of Field Participants." Korean Journal of Journalism & Communication Studies, 67(4), 238-271. 
  74. Parslow, G. R. (2011). "Commentary: How The Internet is Changing The Way We Think, Read and Remember." Biochemistry and Molecular Biology Education, 39(3), 228.
  75. Qadir, J., Islam, M. Q. & Al-Fuqaha, A. (2022). "Toward Accountable Human-Centered AI: Rationale and Promising Directions." Journal of Information, Communication and Ethics in Society, 20(2), 329-342.
  76. Ray, P. P. (2023). "ChatGPT: A Comprehensive Review on Background, Applications, Key Challenges, Bias, Ethics, Limitations and Future Scope." Internet of Things and Cyber-Physical Systems, 3, 121-154. https://doi.org/10.1016/j.iotcps.2023.04.003
  77. Roche, C., Wall, P. J. & Lewis, D. (2023). "Ethics and Diversity in Artificial Intelligence Policies, Strategies and Initiatives." AI and Ethics, 3(4), 1095-1115.
  78. Roe, J. & Perkins, M. (2023). "'What They're not Telling You about ChatGPT': Exploring The Discourse of AI in UK News Media Headlines." Humanities and Social Sciences Communications, 10(1), 1-9.
  79. Ryu, Y. (2022). "Renewing Ethics in The Age of Artificial Intelligence: Focusing on Luciano Floridi's Information Ethics." Communication Theories, 18(2), 5-51. 
  80. Seo, H. (2019). "A Preliminary Discussion on Policy Decision Making of AI in The Fourth Industrial Revolution." Informatization Policy, 26(3), 3-35.  https://doi.org/10.22693/NIAIP.2019.26.3.003
  81. Shim, J. & Chae, S. (2019). "Seeking Possibility of Ethical Issues based on Public Attitude Toward Artificial Intelligence Through Analysis of Social Network Data." The Journal of Humanities and Social Science, 10(3), 1337-1348. 
  82. Short, C. E. & Short, J. C. (2023). "The Artificially Intelligent Entrepreneur: ChatGPT, Prompt Engineering, and Entrepreneurial Rhetoric Creation." Journal of Business Venturing Insights, 19, e00388.
  83. Stahl, B. C. & Eke, D. (2024). "The Ethics of ChatGPT-Exploring The Ethical Issues of an Emerging Technology." International Journal of Information Management, 74, 102700.
  84. Taecharungroj, V. (2023). "'What Can ChatGPT Do?' Analyzing Early Reactions to The Innovative AI Chatbot on Twitter." Big Data and Cognitive Computing, 7(1), 35.
  85. Tredinnick, L. & Laybats, C. (2023). "The Dangers of Generative Artificial Intelligence." Business Information Review, 40(2), 46-48.
  86. Wach, K., Duong, C.D., Ejdys, J., Kazlauskaite, R., Mazurek, G., Korzynski, P., Paliszkiewicz, J. & Ziemba, E. (2023). "The Dark Side of Generative Artificial Intelligence: A Critical Analysis of Controversies and Risks of ChatGPT." Entrepreneurial Business and Economics Review, 11(2), 7-30.
  87. Wong, P. H. (2020). "Cultural Differences as Excuses? Human Rights and Cultural Values in Global Ethics and Governance of AI." Philosophy & Technology, 33(4), 705-715.
  88. Yang, H. (2008). "The Effects of The Opinion and Quality of User Postings on Internet News Readers' Attitude Toward The News Issue." Korean Journal of Journalism & Communication Studies, 52(2), 254-281. 
  89. Yang, J., Yang. S. & Yoon, S. (2023). "An Exploratory Study of Success Factors for Generative AI Services : Utilizing Text Mining and ChatGPT." Information Systems Review, 25(2), 125-144. 
  90. Yoo, N. & Ahn, S. (2023). "A Proposal of Instructional Design Model for Ethical Improvement of Artificial Intelligence." The Korean Association of Computer Education Conference Proceedings, 27(2), 253-256. 
  91. Yoo, S. (2023). "Ethical Issues Posed by 'Generative-AI' (G-AI): Response Strategies for 'Good AI Society'." Journal of The Korean Bioethics Association, 24(1), 1-29.  https://doi.org/10.37305/JKBA.2023.06.24.1.1
  92. Yoo, S., Hyeon, J. & Cho, J. (2023). "Analyzing Public Opinion on News Agency's YouTube Videos and Comments through LDA Topic Modeling." ICT Platform Society Summer Conference Proceedings, 10(1), 50-53. 
  93. Yoon, H. (2021). "Text Classification of YouTube News Media Comments: Focusing on Broadcasting, Cable News, & Conservative Creator Media." Journal of The Korean Data Analysis Society, 23(3), 1399-1416. 
  94. Zhang, C., Zhang, C., Li, C., Qiao, Y., Zheng, S., Dam, S. K., ... & Hong, C. S. (2023). "One Small Step for Generative AI, One Giant Leap for AGI: A Complete Survey on Chatgpt in Aigc Era." arXiv preprint arXiv:2304.06488.