Stop Asking AI to “Write an Essay”
Most students treat ChatGPT like a magic wand and wonder why they get C-grade waffle back. If your prompts are lazy, your results will be too.
Give the machine a specific role
The AI isn’t your tutor; it’s a high-speed intern with no common sense. Stop using generic requests and start assigning specific personas. Tell it to act like a cynical peer reviewer or a ruthless examiner who hates padding. If you don’t define the constraints, it’ll churn out the same beige prose your lecturers are already tired of marking.
Upload your marking criteria
Don’t just ask for feedback. Paste the actual module assessment rubric into the chat and tell the AI to cross-reference your draft against the First-Class criteria. It forces the system to identify specific gaps in your evidence rather than just fixing your spelling. You’re paying nine grand a year; make the tech work as hard as your debt is growing.
The ‘Devil’s Advocate’ protocol
Stop asking for confirmation and start asking for a fight. Tell the AI to find the three weakest points in your argument and tear them apart. It is far better to have a chatbot expose your logic gaps at 2 AM than to let a marker find them three weeks later.
Underground Pro-Tip 💡: Use “few-shot prompting” by feeding the AI three snippets of high-scoring papers from your specific department before you ask for help. It’ll mimic the niche vocabulary and structural quirks of your discipline, making the output look focused rather than generic.
Your degree is a game of strategy, not a typing test—so stop playing it on easy mode and losing.
Leave a Reply