Use and Limitations of Generative AI

When viewing the output of a tool such as ChatGPT, Bing chat or Google Bard it is easy to think it has a level of understanding of the subject being written about. You may also think it is synthesising information in a critical way, paraphrasing and summarising content from multiple sources to build an argument, but that isn't the case.

Technology such as ChatGPT and Google Bard work by building a vast dataset of reference text, sourced from openly available online articles, books and webpages. When responding to a prompt from you, they simply select the statistical "next best word" based on the words and sentences that have gone before – influenced by information you supplied at the start of the chat. In many ways, they work a little like a mobile phone’s predictive text features. 

 There is great potential to use AI tools to support you throughout your education but we must remember there is a big difference between human and artificial intelligence. There is a limit to what AI tools can do although it is not always clear at first glance. AI tools available to us at this moment do not understand the content they generate or what those words or images mean in the real world. 

Keeping in mind the purpose of assessment for and of your learning, will help you use your judgement about the appropriate use of AI in the preparation of your assessed work.

How can AI help your learning?

If you are using AI, think carefully about how you can apply it to support and enhance your learning, and always be transparent about what you have done with the tool and how you have used the content generated.

There are many different ways that you might use AI tools for your learning. They may be valuable in preperation of work particularly at the early stages of planning, thinking and generating ideas, if this use is allowed within the assignment you have been set. You may also find it useful to think about the support AI tools might provide within the writing process. When considering all of these uses however, you must ensure that any work you submit for assessment is your own. 

For example, AI tools might help with:

• answering questions where answers are based on material that can be openly found on the Internet 
• understanding complex materials and ideas in a way that is relatable to you, by providing an alternative voice to learning materials in Canvas or sources on your reading list
• overcoming writer’s block by giving you some inspiration for points to consider
• coherently drafting your ideas around a topic by summarising key points
• planning and structuring your own ideas for written work or presentations
• exploring and comparing different writing styles and approaches intended for a variety of audiences
• reviewing and analysing written materials
• identifying issues in your academic writing style, grammar or structure that you can then address
• debugging code
• generating ideas for the presentation of data, creation of visuals, graphs or images
• summarising notes you have written while reading critically or during lectures 
• quickly transcribing spoken word or describing the contents of images

What are the limitations of AI?

Even though AI generated content is generally well presented and appears convincing, the tools can, and often do, get things wrong. You should always question the output, apply your judgment concerning its reliability, and fact check the information provided. Many AI tools are unable to reference their sources and you will find that citations are often fabricated.

AI is only able to generate responses based on the information it is trained on and the available dataset:

• the data is not always current and tools may be drawing on sources that are months or years out of date
• if the dataset is biased this bias will be transferred into the generated content. Without careful analysis, biases, stereotypes and in some cases Western perspectives may be perpetuated.

AI performs better when it has more sources to draw on. Tools are likely to produce comprehensive outputs on subjects that are widely written on, so if you are looking for assistance with more specialised or niche areas, or cutting edge research, the information may not be so well developed.

AI tools can not apply human critical thinking or the development of an evidenced argument:

• it can synthesise information on a debate but can't assess which side has more strength or credibility
• it may not pick up on subtleties and nuances in the writing, which the writers use to express their stance
• AI tools can't apply the knowledge generated to the real world, except in a very superficial sense

The skills that AI lacks are vital for analysing, developing, and expressing an argument. Critical analysis is a very human skill and often something that is built into the assessments you are tasked with at university. Application of knowledge to differing contexts is also a fundamental skill and your ability to do this helps you to develop higher order thinking skills.