It is hard to imagine dialogue surrounding education in 2023 without the inclusion of artificial intelligence, or AI. What particularly comes to mind are large language models, or LLMs, such as OpenAI’s ChatGPT, which allows users to interact with a chatbot to ask questions and receive answers in a matter of seconds.
Whether in reading responses, open-note tests or even entire essays, AI has become a massive breach in the dam of academic integrity. Although schools have sought to quickly implement devices to catch this covert plagiarism, there is only so much that can be done to plug the leakage of dishonesty AI has brought into the classroom.
As we move forward into the new school year, it is important to acknowledge that it is unreasonable to expect AI tools to disappear. Instead, we feel it would be prudent for academic staff to adapt to teaching alongside AI whilst promoting policies that maintain ethical teaching at UC Berkeley.
One of the largest issues that face this proposal is the nature of coursework. In many classes, students are expected to regurgitate information during assessments and assignments. This layout plays perfectly into AI models like ChatGPT, as it is quite easy to plug in a question that does not involve critical thinking and hopefully receive a factual answer.
Professors and GSIs should model their curriculum around cultivating greater understanding within the student body. Involving less memorization and requiring depth of key knowledge will curb AI’s ability to pervade the academic experience for students.
Additionally, teaching staff should not overcorrect when it comes to altering classroom curriculum. UC Berkeley courses — especially those in the STEM field — have a history of administering exams that essentially set students up for failure.
It is important to make alterations that still allow students to maintain confidence and succeed in courses that are integral to their graduation requirements. Adapting to work with AI should not be a punishment; rather, it should develop the way we perceive technology’s place in the classroom.
From our experience working alongside AI, it can certainly be utilized as a timesaving tool without obstructing one’s ability to absorb knowledge and learn. Whether it is coming up with a quick citation or fact check, eliminating tasks that would otherwise be tedious can pave the way for more time spent unpacking information critical to learning.
As an institution that has often been at the forefront of intellectual achievement, UC Berkeley has already begun developing the way AI is sourced and used. With the creation of the Gorilla LLM and implementing AI tools in its most recent Hackathon, the university has already accepted the fact we must adapt.
It is time for each classroom to follow suit.