Can AI fool us? University Students’ Lack of Ability to Detect ChatGPT
Abstract The evolution that artificial intelligence (AI) has undergone in recent months, especially in its capacity to generate high-quality argumentative texts, has been a disruptive event in academic environments and higher education spaces. One of the current and future significant challenges we face is the difficulty of identifying those texts that simulate our human narrative in a natural language yet have been crafted by an AI. OBJECTIVES: In the present research, we analyze to what extent university students from degrees in Primary Education and Social Education (n=130) can make this distinction. METHODOLOGY: By implementing ad hoc questionnaires, we verify the degree of perception, complexity, and authorship regarding different texts. The texts to be analyzed were various definitions of the concept of education, half made by humans and the other half by an AI that emulated the degree of complexity and expression of the different profiles and human tones. In parallel, the statistical analyses were conducted using the “Advanced Data Analysis” function (formerly “Code Interpreter”) of ChatGPT itself and replicated in SPSS, finding a high similarity between the two, qualitatively consistent in all but one. Additionally, the graphics included were also created using this function. RESULTS: The results indicate the difficulty the students in the sample had in detecting the definitions made by the AI. CONCLUSIONS: Although, as of today, the limits of AI concerning human thought and reasoning are clear, the versatile creative capacity of these language models makes their identification difficult and masks it.
- Referencias
- Cómo citar
- Del mismo autor
- Métricas
Alkaissi, H., & McFarlane, S. I. (2023). Alucinaciones artificiales en ChatGPT: implicaciones en la escritura científica. Cureus, 15(2). https://d66z.short.gy/uQM9Bb
Amir, O., Ariely, D., Cooke, A., Dunning, D., Epley, N., Gneezy, U., Koszegi, B., Lichtenstein, D., Mazar, N., Mullainathan, S., Prelec, D., Shafir, E., & Silva, J. (2005). Psychology, Behavioral Economics, and Public Policy. Marketing Letters 16, 443–454. 10.1007/s11002-005-5904-2
Bulatov, A., Kuratov, Y., Kapushev, Y., & Burtsev, M. S. (2024). Scaling Transformer to 1M tokens and beyond with RMT. arXiv, Article arXiv:2304.11062v2. 10.48550/arXiv.2304.11062
Carrasco, J. P., García, E., Sánchez, D. A., Porter, E., De La Puente, L., Navarro, J., & Cerame, A. (2023). ¿Es capaz “ChatGPT” de aprobar el examen MIR de 2022? Implicaciones de la inteligencia artificial en la educación médica en España. Revista Española de Educación Médica, 4(1). 10.6018/edumed.556511
Carrera Farran, X., & Pérez Garcias, A. (2023). Tecnologías digitales en educación: poniendo el foco en la ética. Edutec, Revista Electrónica De Tecnología Educativa, (83), 1–6. 10.21556/edutec.2023.83.2829
Cortés-Osorio, J. A. (2023). Explorando el potencial de ChatGPT en la escritura científica: ventajas, desafíos y precauciones. Scientia et Technica, 28(01), 3-5. 10.22517/23447214.25303
Crawford, J., Cowling, M., & Allen, K. A. (2023). Leadership is needed for ethical ChatGPT: Character, assessment, and learning using artificial intelligence (AI). Journal of University Teaching and Learning Practice, 20(3). 10.53761/1.20.3.02
Future of Life (2023). Pause Giant AI Experiments: An Open Letter. Future of Life Institute.
Gómez-Cano, C., Sánchez-Castillo, V., & Clavijo, T. A. (2023). Unveiling the Thematic Landscape of Generative Pre-trained Transformer (GPT) Through Bibliometric Analysis, Metaverse Basic and Applied Research, 2, 33. 10.56294/mr202333
González-Geraldo, J. L., Buedo-Martínez, S., & Ortega-López, L. (2023). Los Micromachismos de ChatGPT: Evidencias sexistas a través de UNICODE, implicaciones socioeducativas. XXXV Congreso Internacional de la SIPS. Pedagogía Social en una sociedad digital e hiperconectada: desafíos y propuestas. Salamanca, 4-6 de octubre.
González-Geraldo, J. L., & Ortega-López, L. (2023). Valid but not (too) reliable? discriminating the potential of ChatGPT within higher education. International Conference on Education and New Developments (END 2023). Lisboa, 24-26 de junio.
Gravel, J., D’Amours-Gravel, M., & Osmanlliu, E. (2023, 2023/09/01/). Learning to Fake It: Limited Responses and Fabricated References Provided by ChatGPT for Medical Questions. Mayo Clinic Proceedings: Digital Health, 1(3), 226-234. 10.1016/j.mcpdig.2023.05.004
Hernández-Sampieri, R., & Mendoza, C. (2018). Metodología de la investigación. Las rutas cuantitativa, cualitativa y mixta. McGraw-Hill.
Huang, Y., Wu, R., He, J., & Xiang, Y. (2024). Evaluating ChatGPT-4.0’s data analytic proficiency in epidemiological studies: A comparative analysis with SAS, SPSS, and R. Journal of Global Health, 14. 10.7189/jogh.14.04070
Jover, G., Carabantes, D., & González-Geraldo, J. L. (2023). Asomándonos a la ventana contextual de la Inteligencia Artificial: decálogo de ayuda para la identificación del uso de ChatGPT en textos académicos. Aula Magna 2.0 [Blog]. https://cuedespyd.hypotheses.org/13299
Romero-Rodríguez, J., Ramírez-Montoya, M., Buenestado-Fernández, M., & Lara-Lara, F. (2023). Use of ChatGPT at University as a Tool for Complex Thinking: Students’ Perceived Usefulness. Journal of New Approaches in Educational Research, 12(2), 323-339. 10.7821/naer.2023.7.1458
Sabzalieva, E., & Valentini, A. (2023). ChatGPT e inteligencia artificial en la educación superior: Guía de inicio rápido (ED/HE/IESALC/IP/2023/12). UNESCO e Instituto Internacional de la UNESCO para la Educación Superior en América Latina y el Caribe. https://bit.ly/3oeYm2f
Srivastava, M. (2023). A Day in the Life of ChatGPT as a researcher: Sustainable and Efficient Machine Learning-A Review of Sparsity Techniques and Future Research Directions. OSF Preprints. 10.31219/osf.io/e9p3g
Suárez-Guerrero, C., Gutiérrez-Esteban, P., & Ayuso-Delpuerto, D. (2024). Pedagogía digital. Revisión sistemática del concepto [Digital Pedagogy. Systematic Review of the Concept]. Teoría de la Educación. Revista Interuniversitaria, 36(2), e31721. 10.14201/teri.31721
Taloni, A., Scorcia, V., & Giannaccare, G. (2023). Large Language Model Advanced Data Analysis Abuse to Create a Fake Data Set in Medical Research. JAMA Ophthalmol,141(12), 1174-1175. 10.1001/jamaophthalmol.2023.5162
Tayebi Arasteh, S., Han, T., Lotfinia, M., Kuhl, C., Kather, J. N., Truhn, D., & Nebelung, S. (2024). Large language models streamline automated machine learning for clinical studies. Nature Communications, 15(1), 1603. 10.1038/s41467-024-45879-8
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gómez, A. N., Kaiser, L., & Polosukhin, I. (2017). Attention Is All You Need. Conference on Neural Information Processing Systems (NIPS). USA.
Wang, L., Ge, X., Liu, L., & Hu, G. (2024). Code Interpreter for Bioinformatics: Are We There Yet? Annals of Biomedical Engineering, 52(4), 754-756. 10.1007/s10439-023-03324-9
Amir, O., Ariely, D., Cooke, A., Dunning, D., Epley, N., Gneezy, U., Koszegi, B., Lichtenstein, D., Mazar, N., Mullainathan, S., Prelec, D., Shafir, E., & Silva, J. (2005). Psychology, Behavioral Economics, and Public Policy. Marketing Letters 16, 443–454. 10.1007/s11002-005-5904-2
Bulatov, A., Kuratov, Y., Kapushev, Y., & Burtsev, M. S. (2024). Scaling Transformer to 1M tokens and beyond with RMT. arXiv, Article arXiv:2304.11062v2. 10.48550/arXiv.2304.11062
Carrasco, J. P., García, E., Sánchez, D. A., Porter, E., De La Puente, L., Navarro, J., & Cerame, A. (2023). ¿Es capaz “ChatGPT” de aprobar el examen MIR de 2022? Implicaciones de la inteligencia artificial en la educación médica en España. Revista Española de Educación Médica, 4(1). 10.6018/edumed.556511
Carrera Farran, X., & Pérez Garcias, A. (2023). Tecnologías digitales en educación: poniendo el foco en la ética. Edutec, Revista Electrónica De Tecnología Educativa, (83), 1–6. 10.21556/edutec.2023.83.2829
Cortés-Osorio, J. A. (2023). Explorando el potencial de ChatGPT en la escritura científica: ventajas, desafíos y precauciones. Scientia et Technica, 28(01), 3-5. 10.22517/23447214.25303
Crawford, J., Cowling, M., & Allen, K. A. (2023). Leadership is needed for ethical ChatGPT: Character, assessment, and learning using artificial intelligence (AI). Journal of University Teaching and Learning Practice, 20(3). 10.53761/1.20.3.02
Future of Life (2023). Pause Giant AI Experiments: An Open Letter. Future of Life Institute.
Gómez-Cano, C., Sánchez-Castillo, V., & Clavijo, T. A. (2023). Unveiling the Thematic Landscape of Generative Pre-trained Transformer (GPT) Through Bibliometric Analysis, Metaverse Basic and Applied Research, 2, 33. 10.56294/mr202333
González-Geraldo, J. L., Buedo-Martínez, S., & Ortega-López, L. (2023). Los Micromachismos de ChatGPT: Evidencias sexistas a través de UNICODE, implicaciones socioeducativas. XXXV Congreso Internacional de la SIPS. Pedagogía Social en una sociedad digital e hiperconectada: desafíos y propuestas. Salamanca, 4-6 de octubre.
González-Geraldo, J. L., & Ortega-López, L. (2023). Valid but not (too) reliable? discriminating the potential of ChatGPT within higher education. International Conference on Education and New Developments (END 2023). Lisboa, 24-26 de junio.
Gravel, J., D’Amours-Gravel, M., & Osmanlliu, E. (2023, 2023/09/01/). Learning to Fake It: Limited Responses and Fabricated References Provided by ChatGPT for Medical Questions. Mayo Clinic Proceedings: Digital Health, 1(3), 226-234. 10.1016/j.mcpdig.2023.05.004
Hernández-Sampieri, R., & Mendoza, C. (2018). Metodología de la investigación. Las rutas cuantitativa, cualitativa y mixta. McGraw-Hill.
Huang, Y., Wu, R., He, J., & Xiang, Y. (2024). Evaluating ChatGPT-4.0’s data analytic proficiency in epidemiological studies: A comparative analysis with SAS, SPSS, and R. Journal of Global Health, 14. 10.7189/jogh.14.04070
Jover, G., Carabantes, D., & González-Geraldo, J. L. (2023). Asomándonos a la ventana contextual de la Inteligencia Artificial: decálogo de ayuda para la identificación del uso de ChatGPT en textos académicos. Aula Magna 2.0 [Blog]. https://cuedespyd.hypotheses.org/13299
Romero-Rodríguez, J., Ramírez-Montoya, M., Buenestado-Fernández, M., & Lara-Lara, F. (2023). Use of ChatGPT at University as a Tool for Complex Thinking: Students’ Perceived Usefulness. Journal of New Approaches in Educational Research, 12(2), 323-339. 10.7821/naer.2023.7.1458
Sabzalieva, E., & Valentini, A. (2023). ChatGPT e inteligencia artificial en la educación superior: Guía de inicio rápido (ED/HE/IESALC/IP/2023/12). UNESCO e Instituto Internacional de la UNESCO para la Educación Superior en América Latina y el Caribe. https://bit.ly/3oeYm2f
Srivastava, M. (2023). A Day in the Life of ChatGPT as a researcher: Sustainable and Efficient Machine Learning-A Review of Sparsity Techniques and Future Research Directions. OSF Preprints. 10.31219/osf.io/e9p3g
Suárez-Guerrero, C., Gutiérrez-Esteban, P., & Ayuso-Delpuerto, D. (2024). Pedagogía digital. Revisión sistemática del concepto [Digital Pedagogy. Systematic Review of the Concept]. Teoría de la Educación. Revista Interuniversitaria, 36(2), e31721. 10.14201/teri.31721
Taloni, A., Scorcia, V., & Giannaccare, G. (2023). Large Language Model Advanced Data Analysis Abuse to Create a Fake Data Set in Medical Research. JAMA Ophthalmol,141(12), 1174-1175. 10.1001/jamaophthalmol.2023.5162
Tayebi Arasteh, S., Han, T., Lotfinia, M., Kuhl, C., Kather, J. N., Truhn, D., & Nebelung, S. (2024). Large language models streamline automated machine learning for clinical studies. Nature Communications, 15(1), 1603. 10.1038/s41467-024-45879-8
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gómez, A. N., Kaiser, L., & Polosukhin, I. (2017). Attention Is All You Need. Conference on Neural Information Processing Systems (NIPS). USA.
Wang, L., Ge, X., Liu, L., & Hu, G. (2024). Code Interpreter for Bioinformatics: Are We There Yet? Annals of Biomedical Engineering, 52(4), 754-756. 10.1007/s10439-023-03324-9
González-Geraldo, J. L., & Ortega-López, L. (2024). Can AI fool us? University Students’ Lack of Ability to Detect ChatGPT. Education in The Knowledge Society, 25, e31760. https://doi.org/10.14201/eks.31760
Downloads
Download data is not yet available.
Publication Facts
Metric
This article
Other articles
Peer reviewers
2
2.4
Reviewer profiles N/A
Author statements
Author statements
This article
Other articles
Data availability
N/A
16%
External funding
N/A
32%
Competing interests
N/A
11%
Metric
This journal
Other journals
Articles accepted
22%
33%
Days to publication
230
145
Indexed in
-
—
- Editor & editorial board
- profiles
- Academic society
- N/A
- Publisher
- Ediciones Universidad de Salamanca
To learn about these publication facts, click 
PF is maintained by the Public Knowledge Project
+
−