Services

University Library

We support and advise you in the selection and use of AI tools in the areas of information literacy, literature search, publication advice and research data management. We are able to respond to your individual wishes and find customized solutions.

Contact us:

 

***

 

IDea_Lab

As a research center, the IDea_Lab is dedicated to the topics of digital transformation, artificial intelligence and big data. Our researchers come from various disciplines in the humanities, social sciences, natural sciences and computer science and work in an interdisciplinary manner to understand and shape the digital future. The focus is on the ethics of AI, machine learning methods and data analysis.

Contact: https://idea-lab.uni-graz.at/de

 

 

Help and training

Recommendations on AI in research

Every opportunity should be taken to use AI tools for research purposes as well as administrative activities. The use of the tools can only be learned and refined through practice. Of course, responsible use of the technological possibilities is very important! The following aspects should be taken into account:

  • AI literacy: it is about thoroughly understanding the technology itself but also all AI tools that are used, as well as their capabilities, limitations and T&C(!). AI tools must be carefully selected with regard to the task at hand. It is also important to consider whether the local application of an LLM might be the better approach. You can never improve your prompt engineering skills enough!
    When selecting (generative) AI applications, it may make sense to use commercial products and invest in a paid account if necessary, as this is where the best results can still be achieved.
  • Responsibility: The user is responsible for the results of the AI tools, not the software or the provider. All AI-generated results must be critically scrutinized and reviewed!
  • Transparency: The use of AI tools in research must always be made transparent. It helps to precisely document and disclose the prompts, AI algorithms, data sources and methodologies.
  • Confidential information: No confidential, secret or unpublished data may be transmitted to online AI tools. This especially applies to (peer) reviews and evaluations. If at all, such data can only be processed locally, provided that the software does not send any data to the manufacturing company.
  • Data protection: The possibilities and restrictions arising from data protection guidelines must be observed and respected.
  • Copyright: When dealing with AI tools, only materials that are available for this context (see copyright and licenses) may be used. Open data and open access publications are very helpful here.
  • Publications and funding: AI guidelines from publishers and funding bodies must be observed.
  • Always follow the guidelines for good scientific practice!

Literature, resources

Guidelines from universities