Artificial Intelligence (AI) 

You're probably using AI already!  

Most of us have been using AI (Artificial Intelligence) powered resources for years. Things like YouTube, TikTok and Netflix use AI to suggest new content, and virtual assistants like Amazon’s Alexa use AI to interpret and respond to our questions. You may even already be using AI in your studies, with tools like reference management software or grammar checkers. These kinds of AI are reactive; they provide a predictable output based on the input received. Reactive machines can't function beyond the tasks they were initially designed for, they always respond to identical situations in the same way every time, and they are not able to learn new actions.  

There is no change to the way you can use reactive AI, so you can carry on using your reference management software, spell checkers and other reactive AI tools without having to acknowledge or cite them. Reactive AI doesn't create new content for you.  

So What's New?  

Recent advancements in AI mean we now have generative AI, which includes GPT (Generative Pretrained Transformer) tools like Copilot, Claude, ChatGPT, Bard, perplexity AI and others. These models construct fully formed responses to questions or prompts without ‘knowing’ what the information they’re delivering means.  GPTs are trained on vast banks of data from the internet and use patterns and predictions from these data banks to generate strings of words to simulate human output. Because the output from these tools is generated from patterns it’s often hard to know if the output you’re receiving is correct and reliable Because GPTs aren’t transparent about which information they’re drawing from it is also hard to check or credit the original sources which is problematic for adhering to academic integrity.

Anybody using generative AI in academia needs to understand how it works, be aware of its limitations, use critical thinking in assessing its output and acknowledge or cite use of it in their work.   

UEA policy 

The UEA AI Policy for Teaching and Learning sets out the University’s policy for the use of Generative AI for taught programmes and for taught components of professional doctorates. 

The AI policy for teaching and learning should be read in conjunction with other University policies, especially the UEA plagiarism and Collusion Policy

Please always check school guidance and check with module leaders or supervisors if you are unsure. 

If you are a researcher you should refer to the Generative AI Policy for Research and Innovation. 

The rapid proliferation of generative AI and associated tools is bringing new challenges to university libraries. These challenges are both, assistive - opening new, imaginative and efficient ways of working - as well as disruptive, requiring caution and critical evaluation.

Academic libraries, including the UEA library, are actively engaging with AI in many different ways, but in terms of supporting students, we do the following:

Supporting learning and research

AI tools for learning and research can assist with fundamental library tasks such as search and discovery, synthesis, recommending relevant resources, and data management. 

AI can help with ‘pre-searching’ and generating keywords to enhance discovery.

AI can also enhance search performance in bibliographic, full-text, and other library databases by enabling you to control searches with keywords and filters, for which the AI efficiently processes and delivers relevant results.  

The UEA library provides access to AI features library databases such as SCOPUS AI, ProQuest One Literature, and Statista. This is a more 'controlled' form of generative AI because it is trained on academic databases to predict and generate responses, so we have some reasonable expectation that the results can be verified and cited because this information is provided with the results. 

Promoting AI and “algorithmic” literacy 

Having access to AI tools requires guidance on how to use these tools effectively, ethically, creatively and critically. This includes developing an understanding of the empowering capabilities and benefits of AI, and as well as its complexity, biases, limitations and concerns around privacy and misinformation and how to use it responsibly.

To sensibly use generative AI you will need to have strong skills in strategic searching and critical thinking and a sound understanding of academic integrity. These are the skills that you will be acquiring throughout your studies, and practicing these skills is one the key ways you can gain confidence and ability as a researcher. 

The library offers many resources to support how you discover and use information and scholarly content that is relevant, current, accurate and from a reliable and source that can be attributed, referenced, and verified.

Library databases with AI features

The UEA library provides access to many academic databases. Increasingly the providers are building generative AI as discovery and analysis features. A few of these databases are introduced below. 

Scopus AI 

The Scopus database generally reflects STEM and Social Science fields. Scopus AI uses Generative Pre-trained Transformer (GPT) models that allow you to query the closed and managed set of  journal articles in the Scopus academic databases. The search results in Scopus AI are transparent and clearly explained, with a detailed knowledge graph of works, individuals and concepts to minimise inaccuracy.

Students and staff at UEA can access Scopus AI by logging into the library web pages, searching for Scopus in the A-Z databases list. Once you are on the Scopus home page, look for the AI tab. If you access scopus.com directly from your browser, or if you are off-campus, use the ‘sign in via your organization’ option and search for 'University of East Anglia' to log in.

Note that not all of Scopus is currently covered with the AI feature search. The tool does not learn from full texts, only abstracts and citations, and currently it only indexes back to 2013. The publlisher of Scopus, Elsevier, refers to Scopus AI as a small language model (SLM) as opposed to a large language model (LLM). 

To find out more check out the Scopus libguide from Elsevier. 

Statista 

Statista is a global data and business intelligence platform with an extensive collection of statistics, reports, and insights. It includes credible market data, industry reports, forecasts, opinion polls and infographics. 

Statista now has Generative AI enabled search functions via its 'Research AI' tab. Statista’s Research AI draws all its answers from its datasets using Large Language Model (LLM) technology. You enter your question or 'prompt' into the search box, and the tool produces a brief summary of the information found, followed by links to the databases’ s top 10 relevant data points, with hyperlinks to the actual chart or report in the database. 

While this can generate quick and interesting insights, note that your prompts will not be saved, so you should keep track of your prompts and results by copy-pasting them into a document. In addition, you can't navigate back and forth between the Research AI and Statista search box. So if you click on a link to one of the data sources cited in the results, it will open up the full chart or report, but that will delete out the prompt. 

AI powered research updates from APA PsycInfo

UEA's subscription to American Psychological Association APA PsycInfo database provides access to AI features to optimise your search process. These include visualisation tools for topics and citations, and ways of analytically exploring the PsycInfo database. Staff and students can find out more about the research updates and register for personalized AI powered research updates from APA PsycInfo.

-------

Remember, when using any AI tool: 

• Use the support available from your school, library and learning enhancement team to develop skills in referencing, critical thinking and evaluation.

• Check your assignment brief and school level guidance in case AI is specifically prohibited.

• Keep a log of how you are using AI as you work so that you can credit or cite it where needed when you write up 

• Credit any use of generative AI, this includes searching, gathering concept overviews and formatting, if you used it in any way you must acknowledge it. How you do this will vary from school to school and referencing style. Check your assignment brief and any school guidance or ask your lecturer. 

• Remember- if you would cite a person then cite AI (e.g. if quoting, drawing from AI as the source). If you find yourself citing generative AI and you haven't been explicitly instructed to for your assignment then this could be a sign that you are using generative AI inappropriately (see the red lighted activities on this page). Talk to your module leader if you think this applies to you. 

• Find out about the academic research tools available to you before you decide to use generative AI, AI is just one of your many research and study tools.

Sage Campus free course 'The Art of ChatGPT interactions'

Generative AI tools work by the user inputting prompts (words, phrases, questions, instructions or key words) which the tool then uses to assemble a reply from it's data bank. Well crafted prompts are likely to generate better responses and so its worth taking some time to learn about what prompting invols and ways to get the most out of your prompts. 

Dr Leo S. Lo, Librarian and expert in AI literacy has created a short course called “The Art of ChatGPT Interactions” with the aim to equip learners to develop structured prompt engineering skills, this short and practical course will help you better understand not only how GPTs work but how to get the most out of them.

You don’t need to subscribe to Sage Campus to try it – head over to the demo hub and Access it now for free.