Skip to Main Content

Generative AI for students

About this guide

This guide gives an introduction to generative artificial intelligence (AI) for students at Macquarie University. It

  • Explains some key terms.
  • Provides links to current advice and resources on using generative AI at Macquarie.
  • Lists some common generative AI tools.
  • Suggests strategies for writing effective prompts.
  • Shows tools for evaluating the quality of text outputs.
  • Highlights some ethical issues of generative AI.
  • Gives advice on how to cite (=reference) content created by generative AI.

Its goal is to empower you to make informed decisions and develop your skills in the use of generative AI.

Definitions and key terms

Generative artificial intelligence (AI) is a group of tools that can produce (generate) content, based on natural language input (prompts). For example,  large language models can produce songs in the style of certain singers, short stories in the style of particular writers, or marketing content for a specific audience.

Generative AI tools analyse huge quantities of existing data to detect patterns and features. They then use these patterns to generate new content.

This content generation is driven by statistical probability. For example, Large Language Models  use probability to predict which words should appear in what sequence.

Artificial Intelligence (AI): Computer systems programmed to perform specific tasks, e.g. make predictions based on learned rules. Traditional AI tools perform tasks such as: searching the internet, playing chess, recommending purchases based on behavioural data.

Generative Artificial Intelligence: A type of AI that is used for creating new content based on patterns in the training data. Generative AI tools perform tasks such as: writing text content, creating images, creating music, videos, computer code and much more.

Hallucinations: AIs produce outputs based on prediction. Sometimes this results in a combination of words that equate to something which is not true. The machine cannot "know" anything and as such it cannot differentiate a fact (a truth) from fiction (a falsehood). For this reason, you should never accept information produced by AIs as fact until you have verified it elsewhere.

Large Language Models (LLMs): AIs that generate text content by identifying patterns in existing data and outputting the words most likely to go together.

Natural Language Processing (NLP): The use of computational techniques to interpret and manipulate natural human language. LLMs use NLP.

Go further

Feedback