Skip to Main Content

Generative AI for students

Provides an overview of uses, tools and issues with generative artificial intelligence

What is a prompt?

A prompt is an instruction that you give a generative AI tool, which then produces an output (completion). Writing effective prompts (sometimes called prompt engineering) is a skill that requires practice (and a few strategies) in order for you to get the best results from the tool.

Below is an example of a prompt and output using ChatGPT 3.5:

ACCENT - prompt writing strategies to get you started

The acronym ACCENT outlines six strategies to make your prompts more effective.

  • Alignment: Ensuring the model’s capabilities are in sync with your task.
  • Clarity: Writing clear and concise prompts.
  • Context: Providing relevant context so that the output content is relevant to your needs.
  • Examples: Using examples to demonstrate the format of response that you require.
  • Neutral language: Designing prompts to minimize biases (for example, to minimise assumptions about gender or race).
  • Trial and error: Refining prompts based on the tool's performance and feedback.

This framework for improving prompt engineering was adapted from the 26 guiding principles proposed by Sondos, M. B., Myrzakhan, A., & Shen, Z. (2024):

Alignment: Ensuring the model’s capabilities are in sync with your task.

Check if there another tool that matches your purpose more closely than the well-known large language models.

Your favourite software and platforms (for example, Microsoft Word) are now increasingly adding in AI tools to their functionalities (such as word or sentence completion). However, for some tasks, you might be better using specialised AI tools (e.g. for coding, generating images or searching for academic sources). 

Clarity: Writing clear and concise prompts.

Tip: Make your prompt specific but not too long. Another benefit of being clear is that it can make you think through exactly what kind of output you want),

OK: Help our team to work together effectively on our project.

 

Better: In our class Introduction to Psychology my team (5 people) has to prepare and give a 10-minute group presentation. Help our team identify all the necessary research and writing tasks (SPECIFIC, CONCISE)).

Context: Providing relevant context so that the output content is relevant to your needs.

Tip: Include relevant background information, including constraints (limiters).

OK: Help our team to prepare for our presentation.

 

Better: Our team of five students have to research and present on {topic}. We are not on campus often we so find it hard to meet up (CONSTRAINT). Help us identify and schedule the key tasks, including activities we can complete independently (OPPORTUNITY), along with regular checkins. We have six weeks from today to complete our assessment (CONSTRAINT).

Examples: Including examples of good or bad responses to help explain what you are looking for. This is sometimes called fewshot prompting.

OK: Rewrite my essay for a popular audience.
Better: Turn my essay into a short news article for the student newsletter. Use a conversational style like articles in the previous issue {link}.

 

Including examples to demonstrate the format of response that you require, or the action you want performed. This is also called multi-shot prompting.

Tip: you can upload a document containing the format you wish to use.

OK: Turn the data discussed in this meeting transcript into a table (FORMAT).
Better: Turn the data discussed in this meeting transcript into a table using the same categories as the one in last year's annual report {attached}.

 

Further reading: 

Neutral language: Design your prompts to minimise biases (for example, to minimise assumptions about gender or race).

Tip: Consider whether you need to use she/her or he/him. Often they/them will be sufficient.

OK: An image of a young doctor standing in a hospital ward, holding a stethoscope in his left hand.
Better: An image of a young doctor standing in a hospital ward, holding a stethoscope in their left hand.

 

OK: Suggest ways that waitresses can deal with difficult customers.
Better: Suggest ways that wait staff can deal with difficult customers.

Trial and error: Refining prompts based on the tool's performance and feedback.

Output from the tool can be extremely useful to pinpoint where your prompt is unclear, biased or could benefit from examples.

Output can also show where the tool has weak spots (e.g. providing inaccurate advice, or unable to perform the task).

More ideas for prompting

Rubber duck prompting

Some people say that talking to a rubber duck (which won't ever answer!) can help you find a solution to your own problem simply by talking it through.

In the same way, asking questions (prompts) to an AI tool can help you to think through a problem for yourself. The benefit is not necessarily in the answers it gives you, but how the questions that you ask help you clarify your own thinking.

Further reading:

Asking AI to improve your prompts

Instead of using generative AI tools as search engines (e.g. asking them to provide factual information), try asking them to improve your prompts. 

Other uses include asking for feedback on your writing (provide counterarguments, make your writing more concise).

OK: What are the best strategies for running a project?

 

Better: Suggest three ways I could improve the following prompt: I will be running a project which is due in two months. Help me develop a schedule of work, making good use of the three people involved

 

Further reading:

Prompt chaining

Prompt chaining involves breaking down a task into steps and prompting the AI tool to complete each step, which is used in the following step. Prompt chaining "helps to boost the transparency of your LLM application, increases controllability, and reliability. This means that you can debug problems with model responses much more easily and analyze and improve performance in the different stages that need improvement." (DAIRAI)

Exceptions: This technique will not work when:

  • Your tool does not incorporate outputs (results) from previous prompts. For example, some retrieval augmented generation (RAG) tools such as MultiSearch Research Assistant do not currently support chain of thought prompting.
  • You are using a reasoning model, which already includes these steps its processes. Reasoning models are developed "to think longer and harder about complex tasks, making them effective at strategizing, planning solutions to complex problems, and making decisions based on large volumes of ambiguous information." (OpenAI)

To understand what is possible with your tool, read the tool information carefully or run a few simple tasks before you begin a major task.

Further reading: