Generative AI in Adventist Education:

ChatGPT burst into the public consciousness in November 2022, catching most of us off-guard. This emerging technology can generate essays on many different topics or solve a variety of problems with just minor prompting. For instance, the following prompts to ChatGPT4 generated useful responses in less than 60 seconds:

• Write a 500-word essay on the fall of the Roman Empire.

  https://chat.openai.com/share/5f37237b-2cb2-4930-9af1-319212948aec

• Summarize the history of God’s people as revealed in the Bible.

  https://chat.openai.com/share/23136939-3f08-4fc8-b8f2-90e74d148abd

• Pretend that you’re the Mississippi River. Simulate a dialogue between yourself and your cousin, the Nile River.

  https://chat.openai.com/share/48cceb5e-1e94-441e-ae01-dba37d65ea39

• Write a sample test for a biology module on photosynthesis. Use 15 multiple-choice questions, three fill-in-the-blank, and two essay questions. (Elementary, high school, and graduate level)

  https://chat.openai.com/share/388ea851-2acc-4579-8ab9-06739dc4e9f0

• Provide instructions for wiring and programming an Arduino to simulate a traffic light.

  https://chat.openai.com/share/c7c23983-252f-4c92-914e-3b1f2be5c721

Other related generative AI technologies (such as those listed below) can generate images, presentations, and videos. These emerging resources excite educational technology enthusiasts, as they see the potential of these new tools. They have been tracking artificial intelligence for years, and these thrill seekers are all too eager to exploit the possibilities and the advantages of these emerging technologies. At the same time, these emerging technologies set off alarms in the minds of other educators who see the potential for cheating, plagiarism, or other abuses.

As we begin a new academic year, education faces yet another seismic shift. After navigating the initial years of the COVID pandemic, we anticipated a return to classroom normalcy. However, much like how the virus forced us into Emergency Remote Teaching (ERT), these new, transformative technological advancements now loom, catching educators off-guard just when a return to normalcy seemed possible. The uncertainty surrounding their use may seem threatening at first; however, the more we learn how to use these tools, the more we can maximize their benefit.

ChatGPT is the most well-known of a family of artificial intelligence tools, which are collectively known as “Generative Artificial Intelligence.” “Generative” because its main purpose is to generate creative products from existing data, and “Artificial” because the intelligence it displays is only a simulation of the human creativity that we must continue to foster and grow within the minds of our students and ourselves.

Here are a few examples of generative AI tools that educators and learners alike might want to use:

• ChatGPT (chat.openai.com) – Great for generating text on a variety of subjects. The tool is optimized to be human-like and conversational in tone. However, it is not optimized for accuracy and reliable conclusions. Those who use this tool for information-gathering purposes should validate all findings against their own knowledge or primary sources.
• Perplexity (perplexity.ai) – Similar to ChatGPT; however, this program is great for research and delving deeply into a variety of topics. Highly recommended for academic purposes. Perplexity functions somewhat like a search engine. It suggests answers to your query, provides the primary source for its answer, and provides tips for further investigation.
• SlidesAI.io (slidesai.io) – This is an add-on to Google slides that allows you to generate slide presentations along with graphical backgrounds based on an outline that you provide. It can be a real timesaver.
• Speechify (speechify.com) – This tool allows you to turn written text into speech with a natural and diverse selection of voices.
• Otter (otter.ai) – This meeting assistant can attend Zoom meetings with you. It will listen to the meeting, take notes, and capture slides. Using the information gathered, Otter can provide a summary of the meeting for your review and to distribute to other meeting attendees.
• Grammarly (Grammarly.com) – Yes, that’s right. You’ve been using AI for years and you didn’t even know it. It not only checks spelling on Microsoft Word and Google Docs, but it will also scan for grammar, punctuation, tone, and conciseness.

See https://ditchthattextbook.com/ai-tools/ for a more comprehensive list.

For the sake of clarity and generalization, we will use the generic term “generative AI” for the duration of this article. Before we dive headfirst into the world of AI and examine its potential, we need to explain what generative AI is, what it can do, how it can help, and how it can harm.

Understanding Generative AI

At its core, generative AI functions like an artist who, after studying countless masterpieces, gains the ability to produce original artworks. Similarly, generative AI immerses itself in diverse data, assimilating patterns, nuances, and intricate connections. This knowledge becomes the bedrock for its creative process, allowing it to craft fresh outputs that resonate with human-like quality. This holds immense potential for academia, promising the automation of routine tasks, enhancing research capabilities, and personalizing learning experiences through the intervention of AI tools.

Each member of the academic community must understand that the main purpose of generative AI tools is to be creative and simulate human creativity. Most are not knowledge systems, search engines, or expert systems. Each individual is personally and professionally responsible for all content produced and presented throughout his or her educational and professional endeavors. Additionally, the increasing power and sophistication of generative AI also raise significant ethical concerns and practical challenges, such as the potential for misuse, biased outputs, and privacy violations.

As educators, we understand that generative AI has the potential to transform the world into which our students are emerging. AI tools will become an intrinsic part of the industries and careers for which we are preparing our students. Therefore, we must understand its impact and benefits and strive to incorporate generative AI into our students’ learning experiences. Generative AI can be a powerful tool for personalization, engagement, and feedback. Given the rapid growth and potential impact of generative AI on academic institutions, it is essential to establish guidelines and guardrails that ensure the responsible and ethical use of these technologies.

Implications for Adventist Education: General Assumptions

The following statements reflect the current world in which we find ourselves and are mostly outside of the direct control of our conferences, schools, or educators.

• AI technologies will continue to advance, with transformative effects on various aspects of education, many of which we cannot yet imagine.
• Ensuring access to generative AI tools and resources will be essential in fostering innovation.
• Addressing ethical concerns and establishing guidelines for responsible use will help to uphold the academic institution’s mission and values.

Guiding Principles

• Humans should ALWAYS be at the center of technological advances. AI tools should not be used to replace human knowledge, expertise, or judgment.
• AI tools should be used to support and enhance learning, not replace it.
• Educational administrators must prioritize open dialogue and collaboration about the implications, benefits, and potential risks of generative AI. This effort serves to ensure that these tools are used in a way that is consistent with the philosophy of Adventist education,¹ and the mission, vision, and values promoted by each school or university.
• Strict data protection and privacy standards should be maintained to safeguard individuals’ rights and interests.

The Promise of Generative AI

AI tools can be a valuable tool for enhancing student learning in a number of different ways:

• Exploring or brainstorming topics related to course assignments;
• Enhancing students’ understanding of the curriculum;
• Providing students with opportunities to practice skills;
• Helping students to develop critical thinking and problem-solving skills;
• Critiquing writing and expanding ideas;
• Providing a personalized tutoring environment.

However, it is important to use these tools responsibly and ethically (see Sidebar 1). Teachers and students must educate themselves about the limitations of AI tools and to use them in ways that do not compromise the integrity of the learning experience.

Limitations of Generative AI Tools

The following is a list of current limitations on generative AI tools. These limitations must be understood and acknowledged by learners and educators alike.

• They don’t think or reason. It is important to realize that ChatGPT and its related tools do not possess real intelligence. That’s why it is called artificial intelligence. It uses the Large Language Model (LLM)² to identify patterns and establish links between multiple concepts. This pattern linking allows it to generate text that mirrors human-like writing or conversations. However, this isn’t the product of any personal understanding or creative thought process. Rather, it is a result of complex calculations and probability-based decisions.
• They are dependent on trained content. ChatGPT generative ability, for instance, is drawn from a Large Language Model (LLM) derived from the text it was trained on before September 2021.³ This means it doesn’t have any context for events that have occurred since this date.
• They cannot learn in real time. Because of the current complexity of generating these large language models, substantial time is necessary to process new information. Given the rapid rate at which new data is being produced, these AI tools struggle to stay current.
• They lack contextual understanding. In attempting to answer any question, it is important to understand the context behind it. Unless specifically prompted with contextual understanding, AI tools will generate content based on the context that they have inferred from related content in their data model. This is likely to be only superficially related to the context of a specific query. This is notably problematic when addressing specialized or complex topics, as the content generated might seem relevant superficially but fail to truly engage with the depth of the context. For instance, a generative AI may produce plausible-sounding explanations on medical matters, but without true comprehension, the accuracy and appropriateness of these responses remain questionable.
• They have no opinions, beliefs, or emotions. Especially within the context of Adventism and Adventist education, much of what we promote and teach is based on a specific set of beliefs and a specific understanding of Jesus and the Scriptures.⁴ ChatGPT and other generative AI engines have access to all of the words of the Bible, and all of the writings of Ellen G. White; however, they don’t “believe” these writings. They merely reflect what other people have written about their beliefs. This can be a powerful resource for study and biblical exploration, but not a replacement for true education. “It is the work of true education . . . to train young people to be thinkers and not mere reflectors of other people’s thought. Let students be directed to the sources of truth, to the vast fields opened for research in nature and revelation. Let them contemplate the great facts of duty and destiny, and the mind will expand and strengthen.”⁵
• They have no social, ethical, or moral compass. Although AI responses may reflect certain ethical guidelines, these are determined by design principles and biases of the software designer, not any inherent AI values or consciousness. Generative AI’s seeming creativity and comprehension stem purely from its pattern-recognition capabilities, and not from any true consciousness or intent. None of the current generative AI tools was designed within the context of Seventh-day Adventist beliefs or the values of Adventist education. Consequently, the responsibility falls on both educators and learners to infuse the content produced with our specific belief system. This underscores the vital role humans play in shaping the ethical and moral fabric of the generated material. (See Sidebar 2.)

Conclusion

Generative AI has ushered in a transformative era in Adventist education, presenting a unique blend of opportunities and ethical considerations. While these AI tools contain the potential to revolutionize learning experiences, personalization, and accessibility, their application requires careful navigation. Educators must remember that generative AI serves as a creative augmentation, not a replacement for human knowledge and judgment. By fostering open dialogues, adhering to established guidelines, and prioritizing ethical usage, the educational community can harness the power of generative AI to empower students while upholding the values and integrity of Adventist education. As we navigate this uncharted territory, the responsible integration of AI tools is paramount in shaping a future where technology enhances, rather than supplants, the rich and diverse tapestry of learning.

What’s Next

Let’s share. As Adventist educators, we become stronger by sharing. We plan to have future articles in The Journal of Adventist Education^® (JAE) where different educators will discuss which tools they have used in the classroom. We would like to share what has worked well, and what issues you have found that are left to be addressed.

Acknowledgment by Authors of AI Assistance

• Perused the many articles listed in the reference section of the ChatGPT prompt thread.
• Consulted with colleagues on topics.
• Requested ChatGPT to write the outline and article. (Content produced rejected by authors and not used in this article.) https://chat.openai.com/share/4e77f431-ab3b-4c93-90fb-fdf77e953dec.

  Many of the references included in this thread were either non-existent or unreachable.

• Prompted ChatGPT4 to reflect on the statement: “ChatGPT is an excellent research tool for experts in a given field, but a dangerous tool for the uninformed.” https://chat.openai.com/share/81e5e910-79dd-4bec-9d31-5dc5d3810d8b
• Prompted ChatGPT4: “AI’s limited understanding.” https://chat.openai.com/share/ebb003f9-46af-435c-9031-d7f171230aaa
• Prompted ChatGPT3.5: “What are appropriate uses for ChatGPT in higher education?” https://chat.openai.com/share/826d92dd-6a5c-43c2-a04f-3f7c7818f314.
• Finally, I (D.P.H.) wrote the article myself, based on my reading and understanding.
  • I submitted it to ChatGPT for critique.
  • I asked ChatGPT to help me organize my thoughts better.
  • I asked ChatGPT 4 to use my existing words to rewrite each section.
  • Here is the thread: https://chat.openai.com/share/9c434c82-4f7d-4673-842f-8ba93c7f2ad7.
• We (D.P.H. and F.A.) made liberal use of Microsoft Word’s spell-check “AI” and Grammarly’s style-checking AI.
• So, consider the following: Who wrote this article? David P. Harris, Fred Armstrong, ChatGPT, or all of the above?

This article has been peer reviewed.

David P. Harris

David P. Harris, PhD, is Vice President for Information Systems at Loma Linda University (Loma, Linda, California, U.S.A.), where he dually serves as Associate Professor in the School of Public Health, and Associate Professor of Health Informatics, Information Management, and Administration in the School of Allied Health. He earned a bachelor’s degree in computer science from La Sierra University (then Loma Linda University) and Master’s and doctoral degrees in management of technology in higher education from Claremont Graduate University (Claremont, California). Dr. Harris has many years of service and experience in both the public and Adventist systems in technology management in higher education and as an advisor for learning innovation within the K-12 system.

Fred Armstrong

Fred Armstrong, MA, is Director of Learning Innovation at Loma Linda University (Loma Linda, California, U.S.A.). He is currently completing his EdD in Educational Neuroscience from La Sierra University (Riverside, California), holds a Master of Arts degree in Educational Administration and Supervision from Andrews University (Berrien Springs, Michigan, U.S.A.), and a Bachelor of Music/Elementary Education from Southern Adventist University (Collegedale, Tennessee, U.S.A.). Mr. Armstrong has several years of experience with educational innovation and providing technology support to faculty and administrators. He formerly served as an administrator in the Carolina Conference of Seventh-day Adventists, as principal of the Adventist Christian Academy of Charlotte (Charlotte, North Carolina, U.S.A.), and as the Chair of the Music Department at Forest Lake Academy (Apopka, Florida, U.S.A.).