Education in the Era of Generative AI (2023) PDF

Journal of AI Volume: 7, Issue No: 1, January-December 2023, Pages: 52-62 Education in the Era of Generative Artificial Intelligence (AI): Understanding the Potential Benefits of ChatGPT in Promoting Teaching and Learning David Baidoo-Anu1, Leticia Owusu Ansah2 1Faculty of Education Queen’s Univer...

Journal of AI Volume: 7, Issue No: 1, January-December 2023, Pages: 52-62 Education in the Era of Generative Artificial Intelligence (AI): Understanding the Potential Benefits of ChatGPT in Promoting Teaching and Learning David Baidoo-Anu1, Leticia Owusu Ansah2 1Faculty of Education Queen’s University, Canada. [email protected], 0000-0003-4628-3459 2 University of Cape Coast, Ghana. [email protected], 0009-0004-5141-1392 Recieved: 03.08.2023 Accepted: 11.09.2023 Online: 12.09.2023 Published: 31.12.2023 Review Article Abstract Since its maiden release into the public domain on November 30, 2022, ChatGPT garnered more than one million subscribers within a week. The generative AI tool ⎼ChatGPT took the world by surprise with it sophisticated capacity to carry out remarkably complex tasks. The extraordinary abilities of ChatGPT to perform complex tasks within the field of education has caused mixed feelings among educators, as this advancement in AI seems to revolutionize existing educational praxis. This is an exploratory study that synthesizes recent extant literature to offer some potential benefits and drawbacks of ChatGPT in promoting teaching and learning. Benefits of ChatGPT include but are not limited to promotion of personalized and interactive learning, generating prompts for formative assessment activities that provide ongoing feedback to inform teaching and learning etc. The paper also highlights some inherent limitations in the ChatGPT such as generating wrong information, biases in data training, which may augment existing biases, privacy issues etc. The study offers recommendations on how ChatGPT could be leveraged to maximize teaching and learning. Policy makers, researchers, educators and technology experts could work together and start conversations on how these evolving generative AI tools could be used safely and constructively to improve education and support students’ learning. Keywords: ChatGPT, Education, Generative AI, Teaching and Learning Cite this paper (APA) Baidoo-Anu, D., Owusu Ansah, L. (2023). Education in the Era of Generative Artificial Intelligence (AI): Understanding the Potential Benefits of ChatGPT in Promoting Teaching and Learning. Journal of AI. 7(1), 52-62. 52 Journal of AI 1. INTRODUCTION The 21st century has experienced a rapidly changing landscape in educational practices largely due to advancement in technology (such as artificial intelligence) (Petersen, 2021). Recent progress and expansion in machine learning has led to a more sophisticated innovative technology digital content generation like generative artificial intelligence (AI) (Hu, 2022). Generative modeling artificial intelligence (GAI) is an unsupervised or partially supervised machine learning framework, which generates manmade relics via the use of statistics, probabilities etc (Hu, 2022; Jovanović, 2022). Through advances in deep learning (DL), generative AI creates artificial relics using existing digital content such as but not limited to video, images/graphics, text, audio, video by examining training examples; learning their patterns and distribution (Abukmeil, et al., 2021; Hu, 2022; Jovanović, 2022; Gui, et al., 2021). Extant literature has identified two major generative AI ⎼ Generative Adversarial Network (GAN) and Generative Pre-trained Transformer (GPT) (Abukmeil, et al., 2021; Brown et al., 2020; Hu, 2022; Jovanović, 2022; Gui, et al., 2021). Currently, GAN is the common GAI technique used. GAN uses two neural networks (i.e., generator and discriminator networks). The generator network generates synthetic data (e.g., image of someone’s face), while the discriminator network examines the genuineness of the content to determine whether the content is authentic or not (e.g., whether the image of the human is real or not). This verification process continues until the discriminator network is not able to decipher between the synthetic and real content, and synthetic is recognized as real (Hu, 2022; Jovanović, 2022). CAN is predominantly used for voice generation, graphics and video (Hu, 2022). On the other hand, Generative Pre-trained Transformer (GPT) models use large amount of publicly available digital content data (natural language processing [NLP]) to read and produce human-like text in several languages and can exhibit creativity in writing from a paragraph to a full research article convincingly (or near convincing) on almost any topics (Aydin & Karaarslan, 2023). These models are even able to engage customers in human-like conversation such as customer-service chatbots or fictional characters in video games (Aydin & Karaarslan, 2022; Jovanović, 2022; Korngiebel et al., 2021; Pavlik, 2023). A more sophisticated Generative Pre-trained Transformer (GPT) -3 has recently been developed (Brown, et al., 2020). Using 175 billion parameters, GPT-3 has been developed to enhance task-agnostic and even become competitive with prior state-of-the-art fine-tuning approaches (Brown, et al., 2020). Brown et al., (2020) stated that GPT-3 is ten times more than any previous non-sparse language model. GPT-3 has become the basic NLP engine that runs the recently developed language model ChatGPT, which has attracted attention in various fields, including but not limited to education (Williams, 2023; Tate, 2023), engineer (Qadir, 2022), Journalism (Pavlik, 2023), medical (Nisar, & Aslam, 2023; O’Connor, & ChatGPT, 2023) and economic and finance (Alshater, 2022; Terwiesch, 2023). In the next sections we provided an explanation of what ChatGPT is, and the potential to improve education and students’ learning. We also discussed some limitations and how educators can use ChatGPT to support and improve students’ learning. An earlier version of this work has previously been published in a preprint (see Baidoo-Anu & Owusu-Ansah, 2023). 2. METHODOLOGY We employed an exploratory methodology to understand the potential benefit of ChatGPT in Education. The purpose of this methodology is to systematically review and explore (a) how interactive ChatGPT is, (b) the benefits of ChatGPT and related generative AI in advancing teaching and learning, (c) the possible drawbacks of using ChatGPT and related generative AI in education, reflections and implications for practice. The study was guided by these research questions ⎼(a) How interactive is ChatGPT? (b) What are the potential benefits of ChatGPT and related generative AI in advancing teaching and learning? (c) what are the possible drawbacks of using ChatGPT and related generative ai in education? The study focused on publicly available resources on ChatGPT from November 2022-March 2023. Specifically, we focused on articles published in a peer reviewed journals or preprints as well as newsletter articles and social media. We were also interested in 53 Journal of AI articles which have been published in English Language. 3. WHAT IS CHATGPT? To begin, we asked ChatGPT to explain itself. The prompt we started with was “What is ChatGPT?”. ChatGPT generated this response. “ChatGPT is a large language model trained by OpenAI. It is designed to generate human-like text based on a given prompt or context. It can be used for a variety of natural language processing tasks, such as text completion, conversation generation, and language translation”. (By ChatGPT January 23, 2023). As a follow up, we used another prompt “How does it work?” To know how conversational it is, we intentionally wrote “How does it work?” instead of "How does ChatGPT work?”. It returned with the response below. “ChatGPT is based on a transformer architecture, a type of neural network that has been shown to be very effective for natural language processing tasks. It is trained on a large dataset of text, such as books and articles, and learns to generate text that is similar to the text it was trained on. When given a prompt or context, the model processes the input and generates a response. The response is generated one word at a time, with the model predicting the next word based on the input and the words it has generated so far. The model uses attention mechanisms to focus on the most relevant parts of the input and generate a response that is coherent and appropriate for the given context. After the training, the model can be fine-tuned on a specific task such as question answering or dialogue generation by providing it with task-specific examples and fine-tuning the model on this data. It can also generate text in multiple languages by fine-tuning the model on the multilingual dataset or by providing the model with the language code” (By ChatGPT January 23, 2023). Then the topic of conversation was changed to discuss COVID-19 and impacts on educational deliveries using the prompt “Now discuss how COVID-19 has changed educational pedagogies.” Within seconds it wrote the passage below. The COVID-19 pandemic has had a significant impact on the way education is delivered, with many schools and universities shifting to online or remote learning in order to comply with social distancing guidelines. This has resulted in a rapid shift to digital and online pedagogy, as educators have had to quickly adapt to this new mode of instruction. One of the major changes that has occurred is the increased use of technology for teaching and learning. This has included the use of online learning platforms, such as Zoom, Google Classroom and other video conferencing tools, as well as the use of digital materials, such as e-books, videos, and interactive activities. This has allowed for increased flexibility and accessibility in the delivery of education, as students can now access materials and participate in classes from anywhere with internet access. Additionally, there has been a shift towards more self-directed and asynchronous learning, where students are given more autonomy in their learning and are able to complete coursework at their own pace. This approach allows students to work at a time that is convenient for them, and it can be beneficial for students who may have other responsibilities or who may have difficulty attending live classes. However, the shift to online learning has also highlighted the digital divide, where students in low-income or rural areas may not have access to the technology or internet connectivity needed to fully participate in remote learning. And it also brought challenges such as lack of interaction, lack of motivation and lack of accountability. Overall, COVID-19 has accelerated the adoption of technology and digital pedagogy in education, but it has also highlighted the need for equity and accessibility in the delivery of education. The education system will have to continue to adapt and evolve to meet the needs of students, teachers and parents during these 54 Journal of AI challenging times (By ChatGPT January 23, 2023). Obviously, the outcome clearly shows that ChatGPT is very interactional and capable of not only holding a realistic human-like conversation on a wide range of topics, but also it produces a convincing creative content. Since its maiden release into the public domain on November 30, 2022, ChatGPT garnered more than one million subscribers within a week (Altman, 2022; Mollman, 2022). The ChatGPT took the world by surprise with its sophisticated capacity and became the hottest topic on social media and news outlets including Nature (Stokel-Walker, 2022) and The New York Times (Metz, 2022), among others. The ChatGPT has unprecedented ability to carry out remarkably complex tasks like writing an article (GTP-3 et al., 2021), story, poem, an essay (Lucy & Bamman 2021), ability to provide a summary or expansion of a text, adjusting texts to reflect a different perspective, and even write and debug original computer code (Williams, 2023; Tate et. al, 2023). The extraordinary abilities of ChatGPT to perform complex tasks within the field of education has caused mixed feelings among educators, as this advancement in AI seems to revolutionize existing educational practices. It has become a contentious topic among educators, while some perceive ChatGPT and related generative AI as the future of teaching and learning, as well as educational research. Others are skeptical and view it as a threat and a potential end to most educational activities, which makes teachers and students lazy with no or little analytical abilities. So, the question that begs for an answer is whether ChatGPT and related generative AI are the future for teaching and learning or a threat to the education field? We coauthored with ChatGPT to unpack some benefits of ChatGPT and related generative AI in advancing teaching and learning. We also co-authored to discuss the possible drawback of using ChatGPT and related generative AI in education. Implication for practice was discussed. 4. BENEFITS OF CHATGPT AND RELATED GENERATIVE AI IN ADVANCING TEACHING AND LEARNING We queried ChatGPT to generate the benefit of ChatGPT and related generative AI in advancing teaching and learning. The initial prompt was “What are the benefits of ChatGPT and related generative AI in advancing teaching and learning”. ChatGPT generated very brief (a paragraph) response which was not detailed. Then when we altered the prompt, ChatGPT also altered the response by providing a detailed explanation. The altered prompt was “I want you to discuss with possible reference the benefits of ChatGPT and related generative AI in advancing teaching and learning”. The following detailed responses were generated. ChatGPT, a large language model developed by OpenAI, has the potential to be a powerful tool in the field of education. Here are some specific ways in which ChatGPT can be used to enhance learning and the references that support it: Personalized Tutoring: ChatGPT can be used to provide personalized tutoring and feedback to students based on their individual learning needs and progress. A study by Chen and colleagues (2020) demonstrated that a conversational agent based on a generative model (ChatGPT) could provide personalized math tutoring to students, resulting in improved learning outcomes. The study showed that the conversational agent was able to provide explanations that were tailored to students' misconceptions and was able to adapt to their level of understanding. Automated Essay Grading: ChatGPT can be trained to grade student essays, providing teachers with more time to focus on other aspects of teaching. A study by Kim and colleagues (2019) showed that a generative model (ChatGPT) trained on a dataset of human-graded essays could accurately grade essays written by high school students, with a correlation of 0.86 with human grades. The study showed that the model was able to identify key features of well-written essays and was able to provide feedback that was similar to that provided by human graders. 55 Journal of AI Language Translation: ChatGPT can be used to translate educational materials into different languages, making them more accessible to a wider audience. A study by Johnson and colleagues (2016) demonstrated that a generative model (ChatGPT) trained on a dataset of bilingual sentence pairs could accurately translate between languages, achieving state-of-the-art results on several translation benchmarks. The study showed that the model was able to understand the meaning of sentences in one language and to generate accurate translations in another language. Interactive Learning: ChatGPT can be used to create interactive learning experiences where students can interact with a virtual tutor in a conversational manner. A study by Peng and colleagues (2019) demonstrated that a generative model-based conversational agent could provide effective support for students learning English as a second language, resulting in improved language proficiency. The study showed that the agent was able to understand students' questions and to provide appropriate and relevant responses. Adaptive Learning: ChatGPT can be used to create adaptive learning systems that adjust their teaching methods based on a student's progress and performance. A study by Chiang and colleagues (2021) showed that an adaptive learning system based on a generative model (ChatGPT) could provide more effective support for students learning programming, resulting in improved performance on programming assessments. The study showed that the model was able to understand students' knowledge and to adjust the difficulty of the problems it generated accordingly. Overall, ChatGPT has the potential to be a powerful tool for enhancing teaching and learning by providing personalized tutoring, automated essay grading, language translation, interactive learning and adaptive learning (By ChatGPT January 24, 2023). 5. POSSIBLE DRAWBACK OF USING CHATGPT AND RELATED GENERATIVE AI IN EDUCATION To have an overview of the possible drawback incorporating ChatGPT and related generative AI in education, ChatGPT was asked to discuss with references the possible drawback of using ChatGPT and related generative AI in education. Below is the outcome of my query. While there are many potential benefits of using ChatGPT and other generative AI models in education, there are also some drawbacks to consider. These drawbacks are supported by research studies: Lack of Human Interaction: ChatGPT and other generative models are not capable of providing the same level of human interaction as a real teacher or tutor. This lack of human interaction can be a disadvantage for students who may benefit more from a personal connection with a teacher. A study by D'Mello and colleagues (2014) found that students who interacted with a virtual tutor that mimicked human-like affective behavior had a better learning outcome than those who interacted with a virtual tutor that lacked this behavior. Limited Understanding: Generative models are based on statistical patterns in the data they are trained on, and they do not have a true understanding of the concepts they are helping students learn. This can be a disadvantage when it comes to providing explanations or feedback that are tailored to a student's individual needs and misconceptions. A study by Wang and colleagues (2020) showed that a generative model-based tutoring system lacked the ability to provide explanations that were tailored to students' misconceptions. Bias in Training Data: Generative models are only as good as the data they are trained on, and if the training data contains biases, the model will also be biased. For example, if a model is trained on a dataset of essays that are primarily written by students from a certain demographic, it may not be able to accurately grade essays written by students from other demographics. A study by Bolukbasi and 56 Journal of AI colleagues (2016) showed that a generative model trained on a large corpus of text from the internet exhibited gender bias in its language generation. Lack of Creativity: Generative models can only generate responses based on the patterns in the data they have seen during training, which can limit the creativity and originality of the responses. A study by Ziegler and colleagues (2019) found that a generative model-based music composition system had a limited ability to generate original and diverse melodies. Dependency on Data: Generative models are trained on a large amount of data, and the quality of the model is highly dependent on the quality and quantity of the data. If the data is not sufficient or not relevant, the model will not be able to perform as well. A study by Kocaguneli and colleagues (2019) showed that a generative model-based question answering system performed poorly when the training data was not relevant to the task at hand. Lack of Contextual Understanding: Generative models lack the ability to understand context and situation, which can lead to inappropriate or irrelevant responses. A study by Gao and colleagues (2019) showed that a generative model-based dialogue system had a limited ability to understand and generate contextually appropriate responses in a conversation. Limited ability to personalize instruction: ChatGPT and other generative AI models can provide general information and assistance, but they may not be able to personalize instruction to meet the individual needs of a particular student. (Ribeiro & Vala, 2020) Privacy: There are also concerns about privacy and data security when using ChatGPT and other generative AI models in education. (Ribeiro & Vala, 2020). It is important to keep in mind that ChatGPT and other generative AI models are powerful tools, but they are not a replacement for human teachers and tutors. It is important to use these tools responsibly and in conjunction with human instruction and support (Ribeiro & Vala, 2020). Overall, while generative AI models such as ChatGPT can be powerful tools for enhancing teaching and learning, it is important to be aware of their limitations and to use them in conjunction with other teaching methods that emphasize human interaction and understanding (By ChatGPT January 23, 2023). 6. REFLECTIONS AND IMPLICATIONS FOR PRACTICE Undoubtedly, ChatGPT and other generative AI is already pushing educational boundaries and initiating a significant paradigm shift in existing educational praxis. Since its introduction to the public in 2022, educators have written extensively about potential implications for teachers, students, and policy. Some educators have already started testing the efficiency of ChatGPT by integrating it in their educational activities (e.g., research, teaching, assessment) and found that through automation of certain tasks and processes, ChatGPT is able to save time for other important activities like spending more time with students (Alshater, 2022; Terwiesch, 2023). For example, Terwiesch who is a Professor at the Wharton School of the University of Pennsylvania indicated that it usually takes 20 hours of work to create an exam and another 10 hours for TAs to test the exam and write solutions to it. However, ChatGPT was able to create the exams within 10 hours and reduced TAs time to 5 hours. This shows 100% productivity increase in the “exam writing operation” (Terwiesch, 2023, p. 23). Similarly, Zhai (2022) stated that it took him 2-3 hours to conduct a study on ChatGPT. He said “…the entire process, including generating and trying queries, adding subtitles, and reviewing and organizing the content, took 2-3 hours” (p.9). Researchers were able to ask OpenAI’s GPT-3 to write an academic paper about itself and how it works (Thunstrom, 2022; preprint, GPT et al., 2022). The paper was submitted to an academic journal. Herft (2023) has also identified several ways teachers could use ChatGPT to support and improve their 57 Journal of AI pedagogical and assessment practices. For example, teachers can leverage the capabilities of ChatGPT to create prompts for open-ended questions that align with the learning goals and success criteria of the unit of instruction. Additionally, ChatGPT can be used to also generate quality rubrics that clearly and concisely explain exactly what students need to accomplish to be successful in the various required levels of proficiency. Again, teachers can use ChatGPT to create “prompts for formative assessment activities that provide ongoing feedback to inform teaching and learning” (Herft, 2023, p. 3). Thus, generative AI-powered assessment systems may support the integration of continuous feedback into learning processes by utilising distinctive and atypical artefacts. Students can also use ChatGPT and other chatbots to support their learning. For instance, students could leverage the capacity of these advanced generative AI to provide systematic explanations of certain complex concepts. ChatGPT can serve as a virtual tutor, which can answer students' questions and provide explanations to a wide range of subjects. This can be particularly useful for students who are struggling with a particular topic or who need extra help outside of the classroom. Also, studies have found that non-native speakers of national languages and students with learning and language disabilities (i.e., struggles to write well) will benefit most from these natural language models. There are a wide range of student-centred learning approaches that can be constructed to be played in groups. The ChatGPT and related generative AI have the capacity to create distinct scenarios for students to collaborate to solve problems and achieve goals. In this way, students can learn from each other, which promote a sense of community among leaners. Arguably, ChatGPT has a great potential to support and advance the work of educators, students and researchers. Despite the myriad of potential educational benefits, in its current state ChatGPT has been found to have several serious inherent limitations, such as generating wrong answers and making-up articles that do not exist. For example, an author asked ChatGPT to generate books and articles in a paper he is working on, ChatGPT included a make-up article which does not exist and even provided full bibliographic details of the article with a non-functional URL (Qadir, 2022). These limitations and other glitches have been reported in other studies. Similarly, when we asked ChatGPT to discuss with references the possible drawback of using ChatGPT and related generative AI in education, it fabricated a reference “Ribeiro and Vala, 2020” to support it discussion. When we asked for ChatGPT to reference the citation it stated "A Survey on Generative Artificial Intelligence" by Ribeiro and Vala, which was published in the Proceedings of the 1st International Conference on Emerging Trends in Intelligent Computing and Informatics (ETICI 2020). Deep search for ‘Ribeiro and Vala, 2020’ work revealed that there is no conference presentation by Ribeiro and Vala, 2020 on generative artificial AI. Therefore, we have reasons to believe that this was fabricated article. This confirms a tweet from the CEO of OpenAI, Sam Altman, who described ChatGPT as “incredibly limited, but good enough at some things to create a misleading impression of greatness. It’s a mistake to be relying on it for anything important right now. It's a preview of progress; we have lots of work to do on robustness and truthfulness.” (Tweet on December 11, 2022). Again, a cursory look at the ChatGPT- generated responses in this study reveals that it has no idea of the world after 2021. Hence it could not add any references or information after 2021. This is because ChatGPT was trained with information only up to 2021 (OpenAI, 2022). Given these inherent limitations, educators, researchers, students and other professionals who use ChatGPT and other chatbot should be cautious. 7. CONCLUSION AND THE WAY FORWARD Despite its inherent limitations, it is a nearly undeniable fact that ChatGPT and other generative AI have come to stay and will continue revolutionizing the current educational system. Many have called for ChatGPT to be banned in the schools while others have started developing software to detect AI generated-texts (see, https://writer.com/ai-content-detector/ and https://huggingface.co/roberta-base-openai-detector ). Others have also provided tips that teachers can use to prevent students from using ChatGPT in writing their essays and other school assignments. For example, Elsen-Rooney (2023) reported that the New York City 58 Journal of AI Education Department (NYC) has blocked ChatGPT on school devices and networks so that students and teachers can no longer access ChatGPT. While these various strategies may work for a while, it may not stand the test of time with even more sophisticated generative AI like GTP-5, which is anticipated to come in the not-too-distant future. Currently, extant literature has shown that AI generated-text detectors are not effective with current sophisticated natural processing language models (e.g., Williams, 2023; Tate, 2023). We should not lose sight of the fact that students also have access to these detectors and can alter the text generated to ensure that it becomes undetectable. In lieu of this, it is high time we began to accept the rapidly changing landscape in educational practices and incorporate these changes in our current educational praxis. Moreover, with Microsoft trying to incorporate ChatGPT holistically into its products (Rudolph et al., 2023; Warren, 2023), in no time ChatGPT will be conventional, and it may possibly be too late for educational institutions to rethink their policies and practices to guide and support their students in using ChatGPT safely and constructively. One area that has garnered more attention and become topical is students’ assessment. It is too soon to conclude but very soon educators may need to rethink how students are assessed. They may have to change how assessment is currently done to more innovative assessments. Extant literature has demonstrated that teachers have limited capacity and skills to engage in high quality assessment practices that move learning forward. In lieu of this, educators have consistently called on teachers to develop capacity to engage students in high-quality assessment practices (Earl, 2012; Wiliam, 2011; Willis, Adie, & Klenowski, 2013). Through professional capacity building, teachers could develop the skills needed to harness the power of ChatGPT, and other generative AI to engage in high-quality assessment practices that improve students’ learning. Given the increase in AI even in workspaces, integrating generative AI tools in the classroom and teaching students how to use it constructively and safely could also prepare them to thrive in an AI-dominated work environment after school. Therefore, educators could harness generative AI models like the ChatGPT to support students' learning. Some questions need urgent answers, for example, how can we leverage ChatGPT to support students’ learning? Do we need to train teachers and students on how they can use current generative AI tools to improve teaching and learning? How can we integrate generative AI tools into teacher education programs to prepare teacher candidates or pre-service teachers to effectively use AI tools in their classrooms? Will these generative AI tools close or augment existing digital divide and what is the way forward? Policy makers, researchers, educators and technology experts should work together and start conversations on how these evolving generative AI tools could be used safely and constructively to improve education and support students’ learning. CONFLICT OF INTEREST We have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. DATA AVAILABILITY STATEMENT Data sharing not applicable to this article as no datasets were generated or analysed during the current study. FUNDING Authors received no funding REFERENCES Abukmeil, M., Ferrari, S., Genovese, A., Piuri, V., & Scotti, F. (2021). A survey of unsupervised generative models for exploratory data analysis and representation learning. Acm computing surveys (csur), 54(5), 59 Journal of AI 1-40. https://doi.org/10.1145/3450963. Alshater, M. (2022). Exploring the role of artificial intelligence in enhancing academic performance: A case study of ChatGPT (December 26, 2022). Available at SSRN: https://ssrn.com/abstract=4312358 or http://dx.doi.org/10.2139/ssrn.4312358 Altman, S. (2022, Dec. 4). Twitter. https://twitter.com/sama/status/1599668808285028353?s=20&t=j5ymf1tUeTpeQuJKlWAKaQ. Aydın, Ö., Karaarslan, E. (2022). OpenAI ChatGPT Generated Literature Review: Digital Twin in Healthcare. In Ö. Aydın (Ed.), Emerging Computer Technologies 2 (pp. 22-31). İzmir Akademi Dernegi. Aydın, Ö., & Karaarslan, E. (2023). Is ChatGPT leading generative AI? What is beyond expectations? Academic Platform Journal of Engineering and Smart Systems. Brown, T., Mann, B., Ryder, N., Subbiah, M., Kaplan, J. D., Dhariwal, P., et al. (2020). Language models are few-shot learners. Advances in Neural Information Processing Systems, 33: 1877-1901. Chen, Y., Chen, Y., & Heffernan, N. (2020). Personalized math tutoring with a conversational agent. arXiv preprint arXiv:2012.12121. D'Mello, S., Craig, S., Witherspoon, A., & Graesser, A. (2014). Affective and learning-related dynamics during interactions with an intelligent tutoring system. International Journal of Human-Computer Studies, 72(6), 415-435. Elsen-Rooney, M. (2023). NYC education department blocks ChatGPT on school devices, networks. Retrieved on January 24 2023 from https://ny.chalkbeat.org/2023/1/3/23537987/nyc-schools-ban-chatgptwriting-artificial-intelligence. Earl, L. M. (2012). Assessment as learning: Using classroom assessment to maximize student learning. Corwin Press. Gui, J., Sun, Z., Wen, Y., Tao, D., & Ye, J. (2021). A review on generative adversarial networks: Algorithms, theory, and applications. IEEE Transactions on Knowledge and Data Engineering. doi: 10.1109/TKDE.2021.3130191. Herft, A. (2023). A Teacher's Prompt Guide to ChatGPT aligned with 'What Works Best' Guide. Retrieved on January 23 2023 from https://drive.google.com/file/d/15qAxnUzOwAPwHzoaKBJd8FAgiOZYcIxq/view. Hu, L. (2023). Generative AI and Future. Retrieved on January 23 from https://pub.towardsai.net/generativeai-and-future-c3b1695876f2. Jovanović, M. (2023). Generative Artificial Intelligence: Trends and https://www.computer.org/csdl/magazine/co/2022/10/09903869/1H0G6xvtREk. 0.1109/MC.2022.3192720. Prospects. Johnson, M., Schuster, M., Le, Q., Krikun, M., Wu, Y., Chen, Z., ... & Chen, Y. (2016). Google's neural machine translation system: Bridging the gap between human and machine translation. arXiv pre. 60 Journal of AI Kim, S., Park, J., & Lee, H. (2019). Automated essay scoring using a deep learning model. Journal of Educational Technology Development and Exchange, 2(1), 1-17. Lucy, L., & Bamman, D. (2021, June). Gender and representation bias in GPT-3 generated stories. In Proceedings of the Third Workshop on Narrative Understanding (pp. 48-55). http://dx.doi.org/10.18653/v1/2021.nuse-1.5. Metz, C. (2022, Dec.11). The new chatbots could change the world. Can you trust them? The New York Times. https://www.nytimes.com/2022/12/10/technology/ai-chat-bot-chatgpt.html. Mollman, S. (2022, December 9). ChatGPT has gained 1 million followers in a single week. Here’s why the A.I. chatbot is primed to disrupt search as we know it. Yahoo Finance. https://finance.yahoo.com/news/chatgpt-gained-1-million-followers224523258.html?guccounter=1&guce_referrer=aHR0cHM6Ly9kdWNrZHVja2dvLmNvbS8&gucereferr er_sig=AQAAAFoaVt0Jgmrz33m4F8LQ93tU37YKNp45BNiRU295yW5YSO9Pki rImoDOfHuNU3b9C9E4zFl6W1R6mSZrM5SFK7y0Au9ht wMKUfBuhQ0Y254BzjvjaY ypRR60aUlPgL3xfTqaM81r3SoP4-QEwJD1lW3avB5uVyxoyu0Src7ffpU. OpenAI. (2022). ChatGPT: Optimizing language models for dialogue. OpenAI. Published November 30, 2022. Accessed January 17, 2022. https://openai.com/blog/chatgpt/. O’Connor, S., & ChatGPT. (2023). Open artificial intelligence platforms in nursing education: Tools for academic progress or abuse? Nurse Education in Practice, 66, 103-537. https://doi.org/10.1016/j.nepr.2022.103537. Pavlik, J. V. (2023). Collaborating With ChatGPT: Considering the Implications of Generative Artificial Intelligence for Journalism and Media Education. Journalism & Mass Communication Educator, 0(0). https://doi.org/10.1177/10776958221149577 Petersen, J. (2021). Innovative assessment practices. Retrieved on 2 August 2022 from https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&cad=rja&uact=8&ved=2ahU KEwjhzbGO38T5AhW2hIkEHXPNDdoQFnoECAkQAQ&url=https%3A%2F%2Ffreshgrade.com%2Fwpcontent%2Fuploads%2F2020%2F07%2FFG-Innovative-Assessment-Whitepaper1.pdf&usg=AOvVaw1fWCFBStSE4BqDTXT5_Voi. Qadir, Junaid (2022): Engineering Education in the Era of ChatGPT: Promise and Pitfalls of Generative AI for Education. TechRxiv. Preprint. https://doi.org/10.36227/techrxiv.21789434.v1. Rudolph, J., Tan, S., & Tan, S. (2023). ChatGPT: Bullshit spewer or the end of traditional assessments in higher education?. Journal of Applied Learning and Teaching, 6(1). https://doi.org/10.37074/jalt.2023.6.1.9. Tate, T. P., Doroudi, S., Ritchie, D., Xu, Y., & Uci, m. w. (2023, January 10). Educational Research and AIGenerated Writing: Confronting the Coming Tsunami. https://doi.org/10.35542/osf.io/4mec3. Terwiesch, C. (2023). Would Chat GPT3 get a Wharton MBA? A prediction based on its performance in the operations management course. Mack Institute for Innovation Management at the Wharton School: University of Pennsylvania. 61 Journal of AI Thunstrom, A. O. (2022, June 30). We asked GPT-3 to write an academic paper about itself- then we tried to get it published. Scientific American. https://www.scientificamerican.com/article/we-asked-gpt-3-towrite-an-academic-paper-about-itself-mdash-then-we-tried-to-get-it-published/. Warren, T. (2023, January 9). Microsoft is looking at OpenAI’s GPT for Word, Outlook, and PowerPoint. The Verge, https://www.theverge.com/2023/1/9/23546144/microsoft-openai-word-powerpoint-outlookgpt-integration-rumor Williams, C. (2023). Hype, or the future of learning and teaching? 3 Limits to AI's ability to write student essays. . London School of Economics internet blog. https://kar.kent.ac.uk/99505/. Willis, J., Adie, L., & Klenowski, V. (2013). Conceptualising teachers’ assessment literacies in an era of curriculum and assessment reform. The Australian Educational Researcher, 40, 241-256. https://doi.org/10.1007/s13384-013-0089-9. Wiliam, D. (2011). What is assessment for learning?. Studies in Educational Evaluation, 37(1), 3-14. https://doi.org/10.1016/j.stueduc.2011.03.001. Zhai, X., (2022). ChatGPT User experience: Implications for education. (December 27, 2022). Available at SSRN: https://ssrn.com/abstract=4312418 or http://dx.doi.org/10.2139/ssrn.4312418. Wang, W., Chen, Y., & Heffernan, N. (2020). A generative model-based tutoring system for math word problems. arXiv preprint arXiv:2010.04. 62