Why ChatGPT is such a big deal for education
By Brent A. Anders, Director of Institutional Research and Assessment, Center for Teaching and Learning at the American University of Armenia
Although there have been many different Artificial Intelligence (AI) systems, AI assistants, and chatbots throughout the years (such as the famous/infamous ELIZA AI chatbot from 1966), none of them has been as powerful with such an easy-to-use interface as ChatGPT (Chat Generative Pre-trained Transformer) (Adamopoulou & Moussiades, 2020; Haque et al., 2022). Due to ChatGPT’s ease of use, capabilities in answering virtually any type of question, and the ability to create many different types of content, it has become a major new disruptive technology with some calling it the next Google or a revolution for the Internet (Ritson, 2022; Walsh, 2022). ChatGPT is an important new technology that has ramifications for the world in general and education specifically.
ChatGPT is so easy to use that to interact with it you simply type into a regular text box at https://chat.openai.com/chat, in normal conversational English (no knowledge of computer code required). As of right now (December 2022) it is completely free and fully open to the public (including all of your students). You type in any kind of question or command and within seconds, it will give you an answer or create text content for you in whatever format/style you desire. It can also accept and respond in multiple languages. Its popularity has skyrocketed (5 million users within the first five days of launch on 30 November 2022), and an estimated 20 million users or more have already tried it (Chi et al., 2022). People are excited and interested in using this new AI because it can do so many things.
ChatGPT capabilities are highly impressive in that the system can accomplish many different tasks and processes including the following:
Answer Questions: This system can answer virtually any type of question including college-level math questions (and it will show its work). ChatGPT (based on the GPT 3.5 language model) contains “570GB of data obtained from books, webtexts, Wikipedia, articles and other pieces of writing on the internet. To be even more exact, 300 billion words were fed into the system” (Hughes, 2022, para. 18).
Summarize Information: As an extension to answering questions, ChatGPT can explain and summarize information (anything it already knows about or info you give it) to you in different ways. You can ask it to be more technical in its response or to make it more simplistic. You can even say things like “explain it to me like I’m a 5-year-old,” “explain it to me in a very exciting way,” “explain it to me in an analogy,” or even “explain it to me in a story.” ChatGPT can even use a different tone/voice when explaining/answering/writing content (prompt: explain this as if you were Abraham Lincoln, Carol Dweck, Albert Einstein, Snoop Dogg, etc. – anything you can think of).
Write Text: ChatGPT can write original (will pass plagiarism detection software) works of fact or fiction. It will write out a complete essay and even include a reference section if requested. It can write content in many ways and formats: essays, test/quiz questions, syllabi, poems, song lyrics, movie scripts, calendars of events, bulleted lists, spreadsheets, and much more. This type of language model AI system has already been used to write out entire research articles as well (Adesso, 2022; Frye, 2022; GPT et al., 2022).Feedback: Text can also be given to ChatGPT for feedback. The feedback given is generally very good and has the ability (if properly prompted) to surpass that offered by Grammarly (Kim, 2022). This could serve as a Teaching Assistant (TA) in helping to provide feedback for essays (either used by the instructor or directly by the students).
Write Computer Code: ChatGPT has the ability to create computer code in multiple languages including C++, swift, HTML, Java, and many more. It can dynamically create code or evaluate code that it is given to help identify issues and to improve efficiency.
Limitations and Ethical Considerations
ChatGPT is based on a closed (limited) language model and it does not have access to the Internet; it is limited to information up to the year 2021. This is one reason why its accuracy will be limited. It is also known to sometimes fill in missing information or what has been termed “hallucinate,” meaning expressing things that aren’t 100% accurate (Metz, 2022; OpenAI, 2022). This is why evaluation of created content and critical thinking are as important as ever. The system is also only as good as the information that it has been given. Bias within the information presented by ChatGPT can still exist, so again critical thinking and evaluation are important. Finally, it is also important to remember that this is an AI system that has been pre-trained and is not currently self-training as it goes. It doesn’t understand what it is saying, it isn’t sentient or cognitively aware. It is just answering prompts by recognizing langue patterns based on the language model it has been given and trained on (OpenAI, 2022).
Given that ChatGPT can create complete works (articles/chapters/essays/etc.) ethical considerations need to be taken into account. If a student simply asks ChatGPT to create an essay for him or her and then turns it in is it plagiarism? What if it was turned in as just the rough draft and then the student modifies a large portion of it? What if ChatGPT was only part of the brainstorming and outline, but then provided feedback on the student’s created rough draft? These are questions that academia needs to fully address sooner rather than later, especially as already established AI assistive tools like Grammarly continue to provide more and more “assistance” to students (Grammarly, 2019).
Other things to consider are how to properly attribute an AI in APA format as well as how educational institutions’ student ethics and/or academic integrity policies need to change. At the university I work for it still says “Plagiarism consists of using the words, ideas, concepts or data of another person without proper attribution” (AUA, 2015, para. 6.4.3). In this case, a student could technically argue that if they used an AI like ChatGPT to create a full essay assignment for them it wouldn’t be plagiarism because an AI isn’t a “person” it is a digital tool like a calculator. This is important to address.
Applicability to Education
The power and flexibility of ChatGPT lends itself to many different applications in education. It is important to note that the use of AI in business and industry has reached a point where all students need to gain skills in knowing about and being able to properly use AI (AI Literacy) to be competitive in the jobs market, both locally and globally (Bughin et al., 2018; Cetindamar et al., 2022; Marr, 2022; Meta, 2022; Ng et al., 2022).
Ideas on ChatGPT Use in the Classroom (Anders, 2022; Watkins, 2022)
Prompt Competition: Have students create questions dealing with the class topic and pose the question to ChatGPT, then use predefined criteria to evaluate the AI response (works to develop topic critical thinking).
Reflect and Improve: Pose an important class topic question to ChatGPT in class, assign students (individually or in groups) to copy the response into MS Word, then using Track Changes have them go through and work to enhance the response to make it more correct.
Full Incorporation Option: Present students with a question assignment/essay as usual and allow them to either use ChatGPT or not. If ChatGPT is used, they would need to submit the prompt(s) used and the original ChatGPT output. Students would then submit a Microsoft Word Track Changes document (or Suggesting in Google Docs), showing how they added depth, clarified any misinformation, presented alternative perspectives, and made other improvements to the chatGPT output.
In-Class Preparatory Process: Having students do part of the writing process in class can also be used to ensure students are directly going through an assignment (regarding essay writing). Having students create an initial outline or mind map of what they plan to address would show their direct work.
Maximizing the Localization and Personalization of the Assignment: By requiring students’ responses to incorporate highly localized and/or personalized information, an AI assistant like ChatGPT would most likely only be able to provide general information.
Using More Dynamic Assessment Techniques: Using assessment techniques other than pure writing assignments such as debates, in-class presentations, student videos or podcasts, etc., would require students to provide a more engaging display of their knowledge which would also incorporate soft skills such as presentation/communication capabilities. Students could still use ChatGPT to prepare but would need to perform the assessment requirement on their own and be ready to answer questions on the spot.
Feedback, Scaffolding, and Instructor Assistance: ChatGPT can be used as a digital teaching assistant by providing direct feedback to students. Additionally, students can ask ChatGPT questions either to get clarification of certain class topics or to have everything explained to them again and/or in a different way. ChatGPT can also help instructors in creating content such as making multiple versions of tests, writing student learning assessments, syllabi, rubrics, and more. ChatGPT can also help as a prompt/assignment checker. If ChatGPT can quickly create a passable response to a prompt/assignment, then it should be recrafted to help students more fully use their skills and learning to complete the assignment.
The Future and a Call to Action
ChatGPT represents a major technological shift due to its ease of use (along with free accessibility) and its powerful capabilities. This is not a fad or another simple temporary tech wonder. It is a fundamental move forward in AI capability, applicability, and usability for everyone. Additionally, there are no signs of this development slowing down as many predict that the next version of ChatGPT’s language model (GPT4) will be released by summer 2023 (if not earlier) that will be more powerful by a factor of 100 times or greater (Romero, 2022; Toews, 2022).
AI continues to increase the automation of information presentation to students, which means that faculty’s roles need to shift to focusing more on the development of the student-instructor relationship to maximize relevancy/motivation and comprehension, along with the creation of opportunities for students to actively use and master their learned skills. Faculty should enhance its use of experiential learning through student-centered pedagogy by using instructional techniques such as in-depth classroom discussions, project-based & problem-based learning, simulations/scenarios, and other hands-on learning opportunities.
Faculty need to additionally realize that a great benefit that they provide is through the development of needed soft skills as identified in the Human Skills Matrix (Ouellette et al., 2020) which is comprised of Thinking (such as critical/creative thinking, entrepreneurship, ethics, growth mindset), Interacting (such as communication, collaboration, empathy, negotiation), Managing Ourselves (such as self-awareness, accountability, adaptability, professionalism), and Leading (such as strategic vision, empowerment of others, project management). In this way, faculty will always remain needed and relevant, even as AI continues to develop and become more and more integrated within academia.
PDF: 1 Page Guide to ChatGPT in Education: https://drive.google.com/file/d/1fTtmGz2Cp2nd65mNfQzPyo3beWXc9j9m/view?fbclid=IwAR1ca03NIEONeb8vmgzmkfSRdqKh7UEF0QtCK2i3i1gX1xQ_FoRc3m1StOc
Video: What might ChatGPT mean for higher education part 1: https://youtu.be/Bz7aW6vStBw
Video: What might ChatGPT mean for higher education part 2: https://youtu.be/gB6fM5TMp5c
Adamopoulou, E., & Moussiades, L. (2020, June). An overview of chatbot technology. In IFIP International Conference on Artificial Intelligence Applications and Innovations (pp. 373-383). Springer, Cham.
Adesso, G. (2022). GPT4: The Ultimate Brain. Authorea Preprints. https://d197for5662m48.cloudfront.net/documents/publicationstatus/121528/preprint_pdf/245b0acf06cb7488a14e53a3bcf4d28e.pdf
Anders, B. (2022). ChatGPT (AI) in education - an overview. American University of Armenia. https://drive.google.com/file/d/1fTtmGz2Cp2nd65mNfQzPyo3beWXc9j9m/view?fbclid=IwAR1ca03NIEONeb8vmgzmkfSRdqKh7UEF0QtCK2i3i1gX1xQ_FoRc3m1StOc
AUA. (2022). Student code of ethics. American University of Armenia. https://policies.aua.am/policy/10
Bughin, J., Hazan, E., Lund, S., Dahlström, P., Wiesinger, A., & Subramaniam, A. (2018). Skill shift: Automation and the future of the workforce. McKinsey Global Institute, 1, 3-84.
Cetindamar, D., Kitto, K., Wu, M., Zhang, Y., Abedin, B., & Knight, S. (2022). Explicating AI literacy of employees at digital workplaces. IEEE Transactions on Engineering Management.
Chui, M., Roberts, R., & Yee, L. (2022, December 20). Generative AI is here: How tools like ChatGPT could change your business. Quantum Black AI by McKinsey. https://www.mckinsey.com/capabilities/quantumblack/our-insights/generative-ai-is-here-how-tools-like-chatgpt-could-change-your-business
Frye, B. L. (2022). Should Using an AI Text Generator to Produce Academic Writing Be Plagiarism? Fordham Intellectual Property, Media & Entertainment Law Journal, Forthcoming.
GPT, Thunström, A. O., & Steingrimsson, S. (2022). Can GPT-3 write an academic paper on itself, with minimal human input? HAL Open Science. hal-03701250f.
Grammarly. (2022). How we use AI to enhance your writing | Grammarly spotlight. Grammarly Blog. https://www.grammarly.com/blog/how-grammarly-uses-ai
Haque, M. U., Dharmadasa, I., Sworna, Z. T., Rajapakse, R. N., & Ahmad, H. (2022). I think this is the most disruptive technology: Exploring sentiments of ChatGPT early adopters using Twitter aata. arXiv preprint arXiv:2212.05856.
Hughes, A. (2022, December 20). ChatGPT: Everything you need to know about OpenAI's GPT-3 tool. BBC Science Focus. https://www.sciencefocus.com/future-technology/gpt-3/
Kim, S. (2022, December 20). Replace Grammarly Premium with OpenAI ChatGPT. https://medium.com/geekculture/replace-grammarly-premium-with-openai-chatgpt-320049179c79
Marr, B. (2022). What are the most in-demand AI skills? Forbes. https://www.forbes.com/sites/bernardmarr/2022/06/13/what-are-the-most-in-demand-ai-skills/?sh=9e3ec3d249ca
Meta. (2022). AI and the future of work. Workplace from Meta. https://www.workplace.com/blog/ai-and-the-future-of-work
Metz, K. (2022). Experts warn about the danger of chatbots using ChatGPT and LaMDA to create ‘hallucinations’ without considering what is right. The New York Times. https://www.nytimes.com/2022/12/10/technology/ai-chat-bot-chatgpt.html
Ng, D. T. K., Leung, J. K. L., Su, M. J., Yim, I. H. Y., Qiao, M. S., & Chu, S. K. W. (2022). AI literacy for all. In AI Literacy in K-16 Classrooms (pp. 21-29). Springer, Cham.
Ouellette, K., Clochard-Bossuet, A., Young, S., & Westerman, G. (2020, March). Human Skills: From Conversations to Convergence. MIT J-WEL. https://jwel.mit.edu/assets/document/human-skills-workshop-report
Ritson, M. (2022, December 9). Is ChatGPT the next big threat to Google’s dominance in the AI market? Marketing Weekly. https://www.marketingweek.com/ritson-chatgpt-google-ai
Romero, A. (2022, November 11). GPT-4 rumors from Silicon Valley. The Algorithmic Bridge. https://thealgorithmicbridge.substack.com/p/gpt-4-rumors-from-silicon-valley
Toews, R. (2022, December 20). 10 AI predictions for 2023. Forbes. https://www.forbes.com/sites/robtoews/2022/12/20/10-ai-predictions-for-2023/?sh=d755fa1fab7f
Walsh, T. (2022, December 15). ChatGPT could revolutionize the Internet, but its secrets have experts worried. Science Alert. https://www.sciencealert.com/chatgpt-could-revolutionize-the-internet-but-its-secrets-have-experts-worried
Watkins, R. (2022, December). Update your course syllabus for chatGPT. https://medium.com/@rwatkins_7167/updating-your-course-syllabus-for-chatgpt-965f4b57b00
About the Author
Brent A. Anders, Ph.D., has a doctorate degree in education (focus was in adult learning and online instruction) from Kansas State University, a Master’s degree from the University of Nebraska at Kearney (focus was in instructional technology), and a Bachelor’s degree in Psychology (focus was in human-computer interactions/usability).
Anders has worked in higher education for over twenty years, extensively working as a face-to-face and online instructor, an educational media consultant dealing with video production, live webcasting, student engagement, user experience, instructional technology, and as a course developer/instructional designer.
Anders also served in the U.S. Army for over 26 years, as an airborne infantryman, public affairs reporter, battalion command sergeant major, and as a certified military instructor. Anders retired from the U.S. Army in 2018, after assisting the Kansas State Partnership Program in association with the U.S. Embassy working with the government of Armenia.
Anders has also done multiple public speaking events and presentations throughout the United States and other parts of the world as well as authoring books, research articles, and blog posts dealing with different aspects of instruction, online learning, and educational technology.
His mail is firstname.lastname@example.org.
|Previous page on path||Cover, page 4 of 22||Next page on path|