Stop treating AI like a magic wand
Most students treat ChatGPT like a posh search engine and wonder why they receive C-grade sludge in return. If your prompts are basic, your marks will be too. It is time to stop asking questions and start giving orders that actually move the needle.
Ditch the generic chat bot
If you start a prompt with “Write an essay about,” you have already failed. The AI defaults to a safe, beige middle ground that screams “I did not do the reading” to any lecturer with half a brain. Give it a persona—tell it to act as a cynical Oxford historian or a cut-throat economic analyst—to force it out of its polite, repetitive loops and into something resembling an actual argument.
The marking rubric is your cheat code
Stop guessing what the marker wants. Copy and paste your specific module marking criteria directly into the chat and tell the machine to critique your draft specifically against the “First Class” column. It stops the AI from being “helpful” and turns it into a ruthless editor that knows exactly how to tick the boxes your department demands.
Chain your prompts for actual depth
One-shot prompting is for amateurs who enjoy getting caught by Turnitin. Break the task into granular chunks: ask for a provocative outline, challenge its logic, then demand specific evidence for each point. By forcing the model to iterate, you strip away the tell-tale AI fluff and produce work that looks like it was written by someone who actually bothered to show up to the seminars.
Underground Pro-Tip 💡: Ask the AI to “Identify the three most common academic counter-arguments to this thesis and provide a rebuttal for each using UK-specific case studies.” This adds a layer of critical analysis that usually takes ten hours of library digging, instantly lifting your work above the mediocre masses.
Stop being the AI’s assistant and start being its manager.
Leave a Reply