"Is it cheating if we use AI?" / "Will I get disqualified?"
Many institutions do not have clear guidelines on the use of generative AI in their policies on academic honesty. It is a growing challenge for educators and universities to adapt to the use of AI by students.
Keeping in mind the spirit of assessment, here's how we recommend using AI:
1. Replacing your own opinions -
The essays are meant to understand the student, their thought process and mindset. Directly copy-pasting generative AI responses is not in line with the intent of the evaluation. Always ensure your own personality / thoughts are captured in the answers.
2. Creating false information -
Do not use LLMs to create hypothetical facts about yourself and your past experience.
1. Feedback / second opinions -
It is a missed opportunity if you aren't using AI as a companion to think more clearly. How can I improve my essay? Critique the letter / Review my application etc. Large language models can give insights on your work, which helps you do better.
2. For research -
LLMs save a huge amount of time in retrieving information. For eg:
"Find public universities offering bachelors in digital humanities, with student scholarships, optional test policy and deadlines after 2 months."
"Help me calculate the financial aid!"
"What is the document checklist for Masters in CS at UC Berkeley?"
Universities, professors and students can benefit greatly with custom LLMs.
3. As a writing tool -
Paraphrasing, correcting grammar, improving writing style are valid applications of LLMs. It's like using auto-correct in 2010s. AI just does the writing, but the story still needs to come from you and your past experience.
Q. Will I by caught by AI checkers?
A. AI detectors don't work. OpenAI, the makers of ChatGPT have shut down their AI checking tool because it is not possible to build a reliable system. The AI detection tools that you see in the market are no better than a coin toss - here's an example of the leading AI checker GPTZero classifying the US constitution as "written by AI". Don't worry about these classifiers, rather focus on the guidelines shared by the universities on the fair usage of AI.
