The Big Debate: AI In The Classroom?
As the use of Artificial Intelligence becomes increasingly relevant in society, is it wise for students and educators to implement this technology in their learning?
Have you ever dreamed of letting someone tutor you on how to complete a difficult maths assignment due tomorrow? Do you ever want a magical assistant to help you write a 1000-word essay on photosynthesis? Is there a time you spent hours surfing the internet and textbooks, trying to find the answer to that one niche question assigned by your tutor? Chances are, every student, including you, once harbored such thoughts in your mind. However, unlike previous generations, today’s pupils have a go-to solution to solve all their problems when studying: Artificial Intelligence (AI).
Since the unveiling of ChatGPT on November 30, 2022, the general public has been stunned by the ability of AI to complete complex tasks, solve difficult problems, and generate impressive content within seconds. As discussions over the use and misuse of AI in different sectors of society rage on, some students are implementing the use of AI in their schoolwork and assignments, while educators and school administrators have been cautious over how the technology can be implemented.
This then leads to the question: Should AI be used by students and educators?
What is AI?
Artificial Intelligence is the technology that enables machines and computers to simulate intelligence and problem-solving capabilities typically demonstrated by humans with the brain. When properly trained, AI can complete problem-solving tasks and make critical decisions while continuing to learn and develop.
To know how AI works, it is key to understand how the software operates through an algorithm, a set of rules and instructions machines have to follow to solve a problem. Engineers first input datasets, for example, large amounts of literary text or photos, to the AI for processing through machine learning, which enables the machine to spot and learn patterns. As the machine develops by absorbing more information provided by the dataset, neural networks formed within the AI can start making connections and create predictions. Then, it performs tasks through expert systems that process and distribute content based on the data it has learned.
The concept of AI is not new, the mathematician Alan Turing was famous for developing the first forms of computers and devised the “Turing Test” as a way of examining whether intelligence exhibited by machines is indistinguishable from a human. What is new with the emergence of ChatGPT and image-creating AI generators like Midjourney comes from their nature as generative AI. Unlike previous forms of AI, generative AI can create new text or images traditionally considered to be created by a human. However, software like ChatGPT is classified as “narrow AI,” which can only perform a narrowly defined task or a small set of related tasks. We have not reached the level of “general AI” that can demonstrate intelligent behavior over a wide range of tasks.
How is AI used in education?
Students have become one of the first groups to misuse AI for their purposes. As the use of AI grew more popular in 2023, many students have caught up with the technology by using it with their schoolwork, especially on language arts and social studies assignments. Many high school and university students have showcased how they have used ChatGPT to finish their homework on social media, and there are cases of pupils using the generative AI software to cheat on their exams and create fake essays. The level of ChatGPT being incorporated into language studies has led the experienced English teacher Daniel Herman to write an article titled “The End of High-School English” in the Atlantic magazine.
Educators have struggled to catch up with the alarming trend of plagiarism through AI software, and institutions have pursued different methods to combat the use of AI in schoolwork. Some universities and schools in the US have banned ChatGPT from being used, while the International Baccalaureate Organisation (IB) has vowed not to ban the software but argued the use of AI software “should be in line with the IB's academic integrity policy.” One of the most prominent cases demonstrating how much education institutions have struggled to regulate the use of AI is the University of Hong Kong (HKU). In February 2023, the university temporarily banned the use of AI-based tools in its schoolwork. But months later, HKU decided to reverse its ban on ChatGPT and similar software from being used by students, allowing both staff and students to access the software for free but with a quota.
Why should AI be used by students and educators?
Multiple arguments can be made over allowing AI software like ChatGPT to be used in school. First of all, AI is not going to go away with a ban. Proponents of AI argue that technology will become part of our everyday lives, and students can benefit by using the software to learn how to improve their education. Even if schools block the use of ChatGPT and similar software in assignments, that will not stop students from referencing content created by AI to improve their writing. Technology will gradually be adapted into education, whether parents or educators like it or not. In the past, some have argued against the use of digital devices from being used in schools because of the negative factors. However, many schools have now adopted the use of digital learning as part of their curriculum, and students can be seen in classrooms bringing a laptop or iPad to accompany their studies. Following the trend of devices inside the classroom, AI will inevitably play a role in students’ education.
Secondly, AI can provide a personal dimension to learning. AI can help students learn concepts at their own pace, allowing them to engage with their curriculum in ways a normal classroom setting cannot achieve. Algorithms can decide how to cater to the different needs of students by devising tailored learning schedules and methods, which is great news for disadvantaged students who might be limited in their education due to their learning disabilities like dyslexia or physical disabilities like blindness. By engaging students in their learning, AI increases student motivation and can enhance student performance.
Thirdly, AI can help teachers to make their lives easier. If a student is stuck on a particular question or concept, the algorithm can specifically target the weaknesses in that particular pupil’s understanding, allowing the teacher to spend more time focusing on the larger weak spots in understanding that most classmates share. Students can receive better feedback by AI that teachers don’t notice, such software can also provide useful assistance to individuals if a teacher or caregiver is not present to guide them immediately. In the long run, AI software can evaluate and improve its learning content, and recommend better curriculum and teaching methods to teachers.
Why should AI not be used by students and educators?
On the opposing side, there are also several arguments against the implementation of AI in classrooms just yet. First of all, AI can be easily misused by students to their advantage. As mentioned before, there are numerous cases where students let AI write their homework and essay assignments, and even cheat on their exams. In many cases, even teachers are not able to distinguish a piece of writing between a human author and AI due to overconfidence in their source identification, or potential inexperience reading AI-generated work.
Secondly, even though there are some forms of plagiarism detection software available for educators to distinguish between AI-generated writing and human-created content, these programs can occasionally falsely identify an original work as plagiarism. For example, OpenAI, the company responsible for creating ChatGPT, had to remove its AI classifier tool designed to inform users whether text is AI-generated, citing a “low rate of accuracy.” Independent plagiarism-detecting groups have not fared better. The Washington Post tested a ChatGPT detector from Turnitin which flagged an innocent student’s writing as AI-generated, and there have been cases where human-generated writing has been mistaken for content created by ChatGPT.
Thirdly, there are concerns over potential biases and inaccuracies produced by AI. Any AI software can only be as knowledgeable as the dataset that was provided for its training, and any biases could be amplified through the algorithm. If AI software was trained on biased information, especially on subjects like race and gender, students would get a biased response and potentially perpetuate negative stereotypes of minorities. Datasets could contain outdated information that might not be up to date with the latest developments, which could lead to distrust among students and teachers of the AI software. Not to mention, AI can “hallucinate.” Generative AI chatbots and similar Large Language Models (LLM) can perceive non-existent patterns that humans don’t detect, and result in nonsensical and inaccurate output that might further mislead students.
Verdict
In the end, both sides make valid arguments on how the use of AI could support students, and how flaws designed into the AI could inadvertently affect what information pupils learn from it. Educators need to understand that AI will be a part of daily life in the future, and students need to learn skills relating to how this technology works, and when to spot inconsistencies and errors generated by such software. Technology companies also need to upgrade AI software to ensure fairness and not mistake work by innocent students as AI-generated content.
Teachers play an imperative role in mediating the use of AI in the classroom among students. Despite fears of AI replacing teachers' jobs, many should keep in mind that automation does not replace jobs, but tasks. In the present day, the teacher might be on task of teaching the curriculum to the class and grading fixed assignments set for every student. AI could help teachers lessen the burden of teaching pupils, and point out flaws that they can cover in further lessons.
Should AI be used in classrooms? The answer is probably yes, but we still have a long way to go before that can become a reality.