ChatGPT Goes To School: Teacher's Pet or Nightmare?

Artificial intelligence has penetrated every aspect of our lives, so it’s no surprise that it has also made waves in our education system. One such AI program is ChatGPT, released on November 30, 2022, the AI language model has stirred up quite a bit of controversy as teachers and administrators grapple with the exponential rate at which students are accessing information.

Chat GPT is an AI Chatbot designed by OpenAI to create human-like responses to user chat questions and statements. Highly advanced AI like ChatGPT is not a recent development. Instead, the chatbot's novelty is its user-friendly nature and free online website. Unlike other Chatbots, ChatGPT is capable of remembering past conversations with the user. Where other AI bots are programmed to clear previous questions and responses, ChatGPT builds on them, allowing for more personalized output.

The model's most recent version, GPT-3, has been trained on 3 terabytes, “equivalent to roughly 3 million books’ worth of text,” as explained by the program. And while this may portray the chatbot as all-knowing to most users, others have reported misinformation produced by the program. When asked to recount historical events, calculate financial reports, or asked for medical advice, the chatbot has been known to fabricate information, creating all-too-real responses. Researchers call this "hallucinating" in the tech world.

While students battle misinformation, teachers are met with the developing issue of cheating. To help mitigate and identify plagiarized assignments, ChatGPT has developed a “classifier” for detecting AI-generated responses. This technology works by embedding a unique code or watermark within each question, paper, or exam sheet. When a student takes the test, ChatGPT analyzes their responses and looks for any matches with other answers that contain the same watermark. If a match is found, the student may have cheated by copying someone else's work, allowing for a more accurate and reliable method of identifying similarities between responses. However, the AI classifier has only proven to detect 26% of the AI written responses it has been tested on. In the classroom, this solution simply isn’t effective.

Where students have found ChatGPT useful in getting unwanted work done, so have teachers. Today, teachers who are battling overpopulated class sizes and don't have access to TAs can use ChatGPT to grade the assignments built up on their desks. Proponents of the AI model argue that ChatGPT could generate more detailed feedback than a single teacher could. In the same vein, ChatGPT can assist teachers in their lesson planning. Pulling from a wide range of teaching methods when asked to create a unique activity for the classroom, the chatbot can produce worksheets, projects, and group discussion prompt ideas. However, these advantages of the chatbot call into question the evolving relationship between teacher and student. If neither is performing the tasks specific to their role in a classroom setting, what is the value of our education system when AI is involved?

To combat the misuse of ChatGPT in January of this year, New York City public schools banned the website indefinitely. This catalyzed adversarial relationships between students and the teachers enforcing the ban. The New York public school district realized that the ban was counterproductive, and students were finding ways around it by accessing the program at home and off school wifi. In response, earlier this month the NYC public schools lifted this ban. The chancellor of NYC public schools determined that the lift comes with student well-being in mind. School officials nationwide are coming to the realization that technology’s role in the classroom is inevitable and teaching students to use it responsibly is imperative to maintaining control.

In an educational setting, the rise of AI has created uncertainty for parents, students, teachers, school officials, policymakers, and tech entrepreneurs. Students and teachers alike now have the opportunity to access unoriginal work that can easily be passed off as their own. And now, with that being said, can you guess which paragraph in this post was AI generated?