What schools could do instead of banning ChatGPT

Antonios Karampelas
4 min readJan 8, 2023

--

AI-generated image (Midjourney; text prompt: ban chatbots AI education)

ChatGPT, the chatbot/large language model released by OpenAI in late 2022, is taking education by storm. Currently in K-12, the benefits of using ChatGPT as the teachers’ virtual assistant or as the students’ personalized, virtual tutor, are overshadowed by the potential use of this tool to cheat essays. Indeed, ChatGPT can summarize concepts in simple (or more sophisticated) terms and produce essays fast and effectively.

As this Artificial Intelligence technology is disrupting education, schools need to make important decisions urgently about how to address its risks and use it creatively and pedagogically to improve their students’ learning. Apparently, not all decisions are welcoming this AI breakthrough, as, according to Chalkbeat New York, ChatGPT has been banned from the city’s public schools’ networks, while other districts are also considering the same approach.

This could not be a long-term solution though, not only because students can still have access to ChatGPT when not in school, but also because banning technology might not be the optimal way of using it — as long as there are no imminent safety concerns associated with it. In any case, students will be using powerful, intelligent chatbots in their professional life, therefore schools need to prepare them for the future. So here is what schools could do instead of banning ChatGPT.

Address the root cause of academic dishonesty

Why would the student cheat in the first place? Is it because of low confidence/low self-esteem? Pressure to get good grades? Advancing psychological safety and social-emotional well-being for all students with the support of specialists could do miracles.

Promote Ethos

Morals and ethics are of utmost importance when it comes to discussions about academic honesty. Have K-12 curricula allocated learning space about ethical decision making? Is global, conscious citizenship promoted sufficiently? Harnessing technology for good to oppose potential use that does harm needs to find itself among the big curriculum ideas.

Educate students about AI

K-12 students are going to live an AI-heavy life. Enriching school curricula with learning elements leading to AI literacy is imperative. Students need to understand what AI is and what is not, be able to understand its basic principles, applications, and impact, and be able to interact with it in meaningful and constructive ways. An AI-literate adult is more likely to thrive in an AI-heavy world in comparison to a non-literate one.

Train school administrators and teachers about AI

It is hard to imagine effective long-term decision-making by school leaders and teachers-classroom leaders without the employment of important literacies that are relevant in the Fourth Industrial Revolution, such as AI literacy; data literacy; technology literacy; media literacy. Fear against technology, if it exists, needs to be addressed with care and empathy. Administrators and teachers alike need, and deserve, proper, practical training. Investing in professional development is always worthwhile.

Modernize assessment

Student-centered, diverse assessment schemes can support the minimization of the risk of a students’ grade being affected significantly by a potential misuse of chatbots like ChatGPT. Project-based, inquiry learning that promotes critical thinking, problem-solving, collaboration, creativity, and communication are hard to beat by human-replacing technology. Do students have sufficient opportunities to showcase their learning in meaningful ways throughout the semester/year? Do assessments assist students in climbing the Bloom’s taxonomy ladder? Essays, or single-answer assessment, cannot be the only form of assessment in 2023 and beyond.

Use learning analytics to identify academic dishonesty

The collection and subsequent analysis of longitudinal data of student performance and engagement, among others, could be used to create accurate student profiles that suspicious assessment performance could be compared against. Is an 8th grader’s essay quality (written in English) in contradiction to their MAP scores in English writing up to the 7th grade? Is a student essay’s corresponding lexile level (complexity) not consistent with the typical level for that particular student? Data-based protocols that would activate upon atypical student performance could have students asked to elaborate on their work — been given the benefit of the doubt and always in reference to their learning.

Enjoy the ride

While individuals and societies are increasingly affected by Artificial Intelligence, primary, secondary, and higher education institutions are having the opportunity to design and implement learning experiences that will prepare a whole generation to harness (and live with) technology ethically and, consequently, improve the state of the world. The teaching profession should be recognized for its importance in transitioning to this promising, dynamic future. So why ban?

--

--

Antonios Karampelas

I am a Science, STEAM, and AI educator holding a PhD in Astrophysics. I write about the AI and Learning Analytics paradigm shifts in education.