One interesting unintended consequence of AI might be it role in reshaping examinations in academia and related activities, such as applying for a job.
Multiple surveys indicate that students are rapidly adopting AI in their work. Not surprising. AI tools can be extremely helpful in writing an essay, revising any text, answering questions, drawing, translating, and so forth. Given the potential of these AI tools, how can educational institutions assess the authenticity of a student’s work? What can be done?
First, it reminds me of the early days of Wikipedia. It will be impossible to regulate its use without conducting unacceptable and (I should hope) unlawful surveillance of students. Moreover, it will be valuable for students to be encouraged to use tools that empower them – as AI will. They can use AI in ways that support their learning and communication and be capable of better applying these skills in many careers they may choose to follow. In doing so, however, they should be transparent about exactly how they used AI and for what. Transparency will be key, such as in referencing any source one uses. If you reference and acknowledge the sources you are using as a student or author, why not. That said, you can say that ‘my essay was written by ChatGPT’ but that is about as legitimate as saying ‘my essay was written by my roommate’. Transparent but unacceptable.
Secondly, from my conversations with colleagues still in faculty positions, top universities are already moving away from examinations that would be AI friendly, such as a take-home exam or an essay. They are moving towards more examinations that are done with pen and paper in a monitored setting. Oral examinations are another excellent means. You could use AI and all other means at hand to prepare for such exams, but at the end of the day, you come to an oral or pen and paper examination with your own knowledge and skills.
Oral (viva voce) answers to questions is the way many exams were conducted in the past at Oxford and Cambridge, most often in Latin. The migration of oral to written examinations at Oxford and Cambridge occurred during the 18th and 19th centuries.[1] No one is proposing a move back to Latin, at least in England, but yes, what about more oral examinations in public – in front of your examiners? Doctoral dissertations continue to be most often defended in an oral examination.
Of course, this might resolve the AI problem, but it might mean that students will need to improve their penmanship and oral presentation skills. That can’t be a bad thing. I find my own penmanship suffering from over-reliance on a keyboard. However, it would also raise the cost and time required for examinations. This could mean fewer examinations but with each being more consequential. You can’t save money on making these examinations hybrid, such as online, as a future of AI might even permit an artificial agent being examined – impersonating the student.
I don’t know how these issues will be resolved, but I would encourage any student to start working more with pen and paper. Also hone your oral presentation skills in using proper grammar, facts, and reasoning in your conversations. Preparing for exams might well mean more conversations about the subject matter of your fields.
I doubt that there is a university of note that does not have multiple committees considering these issues. But how would you propose to address the issues raised by AI in the examination of a student’s leaning and intellectual potential in each field of study? Please let me know if you or your committee has solved these issues in a convincing way.
[1]https://www.researchgate.net/publication/248939168_The_Shift_from_Oral_to_Written_Examination_Cambridge_and_Oxford_1700-1900