Online Study Skills Hub: What is Gen AI?

Generative AI is a type of AI that can create new things, like text or images, based on what it has learned. Some examples of Generative AI tools are ChatGPT, DALL-E 2, CoPilot, Google Gemini, Midjourney, Claude, and Perplexity. These tools can write text that sounds human and create pictures from descriptions.

• They can give feedback, like checking your spelling, grammar, and punctuation.
• They can write text and create images.
• They can write, fix, or change computer code.
• They can summarise long documents.
• They can ask and answer questions.

Generative AI tools are trained on lots of information, and they can copy different styles and produce artifacts that seem real. But sometimes, they make things up. This is called "hallucinating." They also sometimes make mistakes or show bias (unfairness). That’s why it’s important to check any information they give you by doing your own research, so you can get the full picture. Don’t rely on AI to do all of this for you.

You should use your own judgement when deciding whether or not to use AI tools. They don’t always make things easier or better. A simple rule is: if you would ask a friend or teacher for help, then it’s okay to ask AI for help too. But if you wouldn’t ask a person for help, then don’t ask AI either.

The university has created clear guidance on how to use Generative AI (Gen AI), like ChatGPT, and will keep improving this as AI changes. These principles apply to a wide range of AI tools and will guide the academic support we offer.

We’re committed to including AI in our teaching and learning in various ways, such as in lectures and seminars, tests, and study support.

• We'll use AI in teaching and assessments to help you learn.
• Everyone will be taught about digital skills, both while you’re at university and after you graduate.
• We’re designing tests that include the use of AI, because you’ll likely need these skills in your job after university.

AI tools are powerful, but they have some important limitations:

• Accuracy: Sometimes, AI can give wrong or confusing information. For example, it can make things up that sound real but aren’t true (this is called a “hallucination”).
• Old Information: Some AI tools are based on old data, so they might use out-of-date facts.
• Bias: AI can sometimes be unfair because it’s trained on data that might favour certain groups of people.
• Understanding: AI doesn’t actually “understand” what it’s saying; it just follows patterns. This means it might struggle with complicated questions or miss things that people would notice.

These limitations show why it’s important to use AI as a helpful tool, but not to rely on it completely. We still need human thinking and judgement.

There are also some ethical (moral) questions to think about as AI continues to develop:

• Be careful not to share personal information about you, fellow students or university staff or any details of placements with AI tools, and understand that these tools may not keep this information safe or private.
• It’s important to keep honesty and integrity when using AI in research.
• Using AI inappropriately could affect your marks, or could mean that you miss out on understanding or gaining skills that you’ll need in your future career.
• There might be issues with copyright when using AI, and we still don’t fully understand all of the rules about this yet.
• We will offer training and support to help you use AI safely and responsibly.