By Pachara Aungsusuknarumol

Last month, ChatGPT, a new form of artificial intelligence (AI) software, took an exam at the Wharton School of the University of Pennsylvania and passed with a B grade. This has sparked discussions among educators and policymakers about whether AI should be banned from school due to the fear of assignment cheating, and if not, to what extent students should be allowed to utilize the AI technology. Regardless, ChatGPT continues to prove as a useful tool for internet users with high potential for educational purposes.

As a graduate student of International Affairs, I found ChatGPT useful for sparking creative ideas, especially in writing. I use it to proofread and check grammar, asking the bot to "check the grammar of this sentence" for quick feedback. I have employed this practice for various tasks, like emails, cover letters and essay titles. For instance, when I asked ChatGPT to "create a catchy article title for an op-ed arguing ChatGPT should be used in the classroom," it suggested "Bringing Technology to the Classroom: Embracing ChatGPT to Enhance Student Learning." Although it wasn't exactly what I was hoping for, it gave me a solid starting point.

Interactions between students and ChatGPT, particularly those using English as their second language, can be beneficial. Students can learn from their writing mistakes through direct feedback from ChatGPT, which is both quick and free. It is reminiscent of a traditional method I encountered in 9th grade where we were tasked to write letters to foreign pen pals online to practice English. Those interactions with native speakers helped me learn English, and now ChatGPT offers the same experience but with instant responses. In fact, this technology can help students become more independent learners, as they have instant access to information and can ask questions without being limited by the knowledge of their teachers or peers.

Compared to other writing tools used by college students such as Grammarly, ChatGPT is much more versatile. While Grammarly excels in detecting and fixing typical writing errors, it lacks the capacity to generate new text based on user prompts. ChatGPT, on the other hand, can not only suggest text but also respond to requests such as "make this more concise" or "summarize." Working with ChatGPT is like an interactive gaming experience for writing.

However, the advanced capabilities of ChatGPT come with potential risks and challenges. Students might be tempted to use it as a crutch for their own learning, thereby sacrificing the development of critical thinking and writing skills. In the worst-case scenario, the use of AI technology in an academic setting raises ethical concerns about academic integrity and the potential for plagiarism. Considering these concerns, the New York City education department has decided to block ChatGPT's access from school Wi-Fi and prohibit its use in helping students with their homework assignments.

While it is true that ChatGPT can generate its own words and may be used to complete university assignments, it is essential to recognize that there are limitations to what ChatGPT can do. For instance, it is impossible for ChatGPT to write a midterm paper for a Conceptual Foundation Course discussing realism, liberalism and constructivism in-depth. While the software can provide an overview of each ideology based on information available online, it is unable to scrutinize and debate the intricacies of the theories to the extent required in the course, where students are expected to construct their arguments using over 30 specifically assigned readings.

Additionally, ChatGPT would not be successful in completing an assignment for an International Security Policy class, where a portion of the work requires identifying the student's understanding of assigned readings and properly citing them. The software may be able to handle simple university assignments, but it is not capable of producing a research paper or academic essay with proper citations. Its capabilities fall short of this task. As a result, it would be nearly impossible to rely solely on ChatGPT to complete academic assignments.

Does all this sound like AI is a plagiarist’s best friend? Fortunately, the poacher can turn gamekeeper, as AI can also be used to root out cheating. Software such as GPTZero is currently being developed to evaluate sentence complexity and determine whether the input text is more likely to have been generated by a human or an AI. This technology could complement the use of plagiarism detection programs like Turnitin, which are widely used by universities to discern instances of academic dishonesty and plagiarism.

Rather than banning the technology, schools and universities should embrace it and educate students on its limitations and ethical considerations. My personal resolution to these issues is to use ChatGPT to enhance my work, not to get ChatGPT to do it. This approach enables a clear understanding of how ChatGPT can be used, while still fostering the development of critical thinking skills through assignments and the use of AI.

Pachara (Pat) Aungsusuknarumol (MIA '24) is a Fulbright Scholar from Thailand studying International Security Policy (Cybersecurity Track) with a specialization in Technology, Media, and Communications (TMaC).