When considering the role of AI in education, it's important to consider both the potential benefits and risks to education. As with any innovation, there are both incredible opportunities for learning and potential challenges and potentially problematic situations that may arise.
Any discussion of tools like ChatGPT or other AI systems in education, should acknowledge the potential benefit that AI systems bring to education, such as the ability for students to explore concepts on their own with the chatbot, ahead of discussions with their peers or instructors. While it's true that ChatGPT is not always factual and does 'hallucinate' quite often, it can also be a powerful tool for engaging students and allowing them to explore topics in a somewhat safe environment, as ChatGPT has recently improved its response moderation to exclude sensitive or offensive topics to the best of its ability. When not in the classroom, ChatGPT provides the same kind of engagement for students, to be able to explore topics and gain additional thoughts and feedback on their work or their assignments with an interactive content source. This can be especially powerful as an equalizer for students that don't have a home environment where space for academic conversations are not provided, providing a space, even though it is just digital, where academic subjects and conversations can be had and explored.
Anecdotally, one can imagine a classroom buzzing with enthusiasm as students interact with ChatGPT, asking complex questions and receiving immediate responses. This kind of personalized and interactive learning experience can foster curiosity, ignite a passion for knowledge, and empower students to become active participants in their own education. By leveraging AI, we can create learning environments that adapt to individual needs and promote deeper engagement with the material.
However, while AI has the potential to level the playing field and provide equal access to educational resources, it can also exacerbate existing inequalities. One concern is the impact that use of AI will have on student performance. Because not all students have equal access to technology or the necessary digital literacy skills, disparities can emerge in not only who has access to AI systems, but who learns the most and benefits the most from access to AI. Over time, even small differences in equity can define major life-paths for students.
Of course, one of the largest concerns that many educators are discussing is the use of ChatGPT to completely plagiarize or complete assignments. While conventional wisdom has long suggested that the type of students that do this plagiarism are struggling students, data that we've discussed in other blog posts, shows that often its only the struggling students that are getting caught for this kind of plagiarism, while the top performing students are able to leverage AI to create an even larger advantage for themselves.
Unfortunately, tools like Turnitin and GPTZero that claim to have up to 98% accuracy in detecting AI, have efficacy that is far from what is advertised. Interested educators can try for themselves, as I did while writing this blog post. As a test for this article, I told ChatGPT to write an essay on the importance of Pens in education, then directly pasted it, with no editing, into one of the myriad 'AI Detector' tools that are currently online. The screenshot of the result is below. Despite copying and pasting directly from ChatGPT, there is apparently only a 7% likelihood that my text was written by AI!
The lack of efficacy of all of these types of 'detector' tools not only allow students to rampantly plagiarize content, but also falsely flag human-written content as AI generated. Savvy educators are beginning to realize that even a 90% likelihood of something being written by AI, is not enough to make an academic determination against a student for plagiarism.
So, while detector tools are popular now, it's very likely that in the coming year or two, it will become clear to educators that these tools are unreliable and ultimately have the potential to do more harm than good.
This issue is why we created Authoriginal - a 'prevention' based tool that completely flags all forms of plagiarism, from ChatGPT, to Jasper.ai, to copying from a friend's essay, all with a simple LTI integrated tool that is just as easy-to-use as EdTech giants like Turnitin. The face of plagiarism is changing, with the vast majority now being conducted through AI. If you're interested in seeing more about how Authoriginal works, reach out to us at sales@authoriginal.com, and we'd love to set up a demo with you.
Get a Demo of Authoriginal
