ChatGPT Returns as Kids Go Back to School

ChatGPT Returns as Kids Go Back to School

Embracing the Power of Generative AI in Education: Challenges, Opportunities, and Unplanned Lessons

OpenAI ChatGPT

Last winter, the unveiling of OpenAI’s incredibly sophisticated chatbot, ChatGPT, caused quite a stir in the education community. Many educators feared that generative AI like ChatGPT would lead to rampant cheating and plagiarism, rendering traditional English classes obsolete. Some schools even went as far as banning the use of ChatGPT on their networks. However, as we enter a new school year, it seems that early alarm has given way to pragmatism, as educators are starting to see the potential benefits and challenges of incorporating generative AI into the classroom.

One of the main concerns surrounding generative AI is its tendency to “hallucinate” or fabricate information. Students have quickly caught on to this, and educators have recognized the importance of assignments that require critical thinking skills. Instead of viewing AI as a threat to traditional teaching methods, many teachers are using it as a tool to spark conversations and new ideas in the classroom. There is a growing consensus among educators that relying solely on tools to catch AI-generated cheats may not be effective, and that a more nuanced approach is required.

Lisa Parry, a K-12 school principal and AP English Language and Composition teacher, is cautiously embracing generative AI. While she still worries about the potential for cheating, she also acknowledges that plagiarism has always been a concern, and sees the value in using ChatGPT as a powerful search engine to help students brainstorm essay topics. Parry recognizes that generative AI has the power to both enhance and undermine academic progress, and believes it’s important not to dismiss its potential entirely.

The debate around generative AI in education goes beyond concerns about cheating. Educators are now grappling with the fact that this technology has the potential to radically reshape their jobs and the world in which their students will grow up. There is a sense of unease in the education community, as they navigate a technology that they didn’t ask for and try to determine its place in the classroom. Lalitha Vasudevan, a professor of technology and education, sums up this sentiment: “We are taught different technologies as they emerge, but we actually have no idea how they’re going to play out.”

Cheat Codes

Turnitin AI Detection Tool

In the race to detect AI-generated work, tools like Turnitin’s AI detection tool have emerged. These tools analyze written submissions and highlight portions that may have been generated by AI. Between April and July, Turnitin reviewed over 65 million submissions and found that 10.3 percent contained AI writing in potentially more than 20% of the work. However, these systems are not foolproof, with a 4% false positive rate in determining AI-generated sentences.

The limitations of AI detection tools, like Turnitin’s, mirror the limitations of generative AI itself. Both rely on training data that can introduce biases, potentially flagging certain writing styles or vocabularies as AI-generated more frequently. There is a concern that these tools may disproportionately flag certain students, leading to false accusations. The detection tools still have a long way to go in terms of accuracy and fairness.

Given the limitations and potential biases of AI detection tools, some schools are pushing back against their use. The University of Pittsburgh’s Teaching Center and Vanderbilt University, for example, have chosen not to endorse or utilize AI detection tools due to reliability concerns. Additionally, even OpenAI, the creator of ChatGPT, has shut down its AI Classifier tool, citing low accuracy in determining the origin of text. It is clear that the education community is still seeking better ways to handle AI in the classroom.

The New School

AI-Assisted Learning

Generative AI has the phenomenal ability to regurgitate vast amounts of information from the internet, but it lacks critical thinking skills. Some teachers are using this as an opportunity to design lesson plans that challenge AI’s capabilities. For instance, educators may use a chatbot to generate work and then assess whether the quality meets their expectations. If the chatbot easily produces decent work, it could indicate that the assignment needs adjustment and greater complexity.

This cat-and-mouse game with generative AI is reminiscent of challenges posed by copying from books or the internet. Emily Isaacs, from Montclair State University, believes it is the task of educators to persuade students that learning itself is worthwhile, and AI is simply a tool to aid their educational journey. David Joyner, a professor at the Georgia Institute of Technology, adds an AI chatbot policy to his syllabus, encouraging its use as a learning tool rather than a replacement for learning.

The preparation for a world shaped by AI also extends to middle school teachers. Theresa Robertson, a STEM teacher in Kansas City, plans to engage her students in discussions about AI and its potential impacts. She believes it is crucial to expose kids to AI and guide them in understanding the ethical aspects, preparing them for a future where AI is increasingly pervasive.

Unplanned Lessons

AI in the Classroom

There is currently no consensus or “best practice” for teaching in a post-ChatGPT world. Guidance for teachers in the United States remains scattershot, with school districts making their own decisions on whether to allow ChatGPT in classrooms. Larger school districts have diverse stances, varying from blocking access to unabated use. However, outright banning ChatGPT can have unintended consequences, exacerbating the digital divide and denying students, particularly those from low-income and marginalized communities, the opportunity to learn from this technology.

Although incorporating generative AI into lesson planning may seem overwhelming to teachers already recovering from the disruptions caused by the pandemic, there is still a need for a balanced approach. Noemi Waight, an associate professor of science education, highlights the importance of considering the equitable and justice-oriented aspects of AI in education. Banning AI tools without proper analysis could perpetuate existing inequalities in access to technology.

On the other hand, some educators have found unexpected benefits in using generative AI. Bill Selak, the director of technology at the Hillbrook School, began using ChatGPT to generate prompts for an AI image generator after a tragic school shooting. The generative AI not only aided students in processing complex topics like racism and climate change but also amplified human creativity in unexpected ways.

In this rapidly evolving landscape of education and AI, there are still many unanswered questions. While teachers grapple with the challenges and opportunities of generative AI, it is important to keep exploring new ways to integrate AI as a learning tool. By embracing AI and nurturing its ethical use, educators can prepare students to navigate an increasingly AI-driven world.

Disclaimer: The article and all imagery are fictional and created for illustrative purposes only.