Use AI as inspiration - not as a source!
The goal of studying is to think for yourself and learn new skills that can later be applied in work and everyday life. Consider thus carefully whether you choose artificial intelligence or intelligence augmentation. By using only an AI application to produce text, you might get the job done, but you do not learn anything. Learning becomes possible when you use these applications to augment your intelligence, i.e. for brainstorming, reflecting and analyzing.
AI applications using generative AI (GenAI) are able to produce different types of content, for example, text, images and videos, because their responses are based on large training data. One type of generative AI are large language models which produce human-like language. The responses of language models are based on the probabilities of words' occurrence. The answer generated by an AI application may thus sound plausible, even if it is based on incorrect information. This phenomenon is called hallucination where AI presents a response as factual even though it does not know the answer. The response might also be biased because the training data entails human biases.
Search-based generative AI, on the other hand, searches for information also from external sources, after which it produces a response based on them. Because the answer is based on information retrieved from different sources, it may be more accurate than information generated by generative AI alone. Responses may still be false or misleading, because AI might use the sources it finds incorrectly.
AI literacy, which means the ability to understand, use, evaluate and apply AI effectively, responsibly, ethically and sustainably, is essential in studies, workplace and everyday life. AI literacy skills should be maintained and developed in order to be able to evaluate information, for example, to identify biases, opinions and deep fakes.
Be critical and check the accuracy of the answers of AI apps from other sources. AI provides powerful tools for information searching, but clarity, transparency and responsibility must be kept in mind when utilizing them.
AI applications can be used in information searching to, for example, familiarize oneself with the search topic and plan the search. You can brainstorm search terms describing your topic and get help with creating search statements. However, databases might have slightly different practices when it comes to search techniques, so the search statements generated by AI may not work effectively in every database. Many scientific databases today contain AI assistants and functions that make information retrieval easier.
AI is most useful when the topic is familiar enough for the user to be able to evaluate the text and information it produces. It is also good to know how various search techniques work and the purpose of subject terms and keywords.
Prompting is the process of giving instructions to an AI application to get specific answers. Prompt can be, for example, a question, instruction or description. A good prompt is a clear statement in which you explain what you want the response to contain. With more specific follow-up questions, you steer the search in the right direction.
You can ask AI applications to for example:
Artificial intelligence has its own limitations and no real knowledge or understanding. Therefore, content produced by AI should always be viewed critically, because it might be incomplete or even incorrect. If an AI application does not know the answer, it gives the most likely option based on the data it is trained on.
Although different AI applications can be used in your own language, they usually work in English, so both your question and the response you receive are translations. This is worth considering when assessing the correctness of the response.
The user always has the moral and legal responsibility for the response AI gives, so do not blindly rely on AI generated content but check that the information is correct and the sources provided real. Remember to be critical: where does the information come from, is it reliable?
When using an AI application, you are not only responsible for the correctness and use of the text it produces but also for the information you give it. You must take into account the copyrights of the material, as well as the ethical aspects related to the use of AI. In addition, it is good to be aware of the effects on the environment.
Presenting text produced by someone else as one's own is fraudulent activity, whether it is produced by a friend, researcher or language model. Before using an AI application to produce text or do tasks for you, consider the consequences carefully. Also check the AI guidelines for each course. Fraud can have serious consequences, even if the use of an AI application might not seem wrong. Make wise decisions and take responsibility for your academic work.
Material made by another person should not be fed to AI without permission. Also licensed materials, such as scientific articles, may not be given to AI applications. CC licenses for articles do not override this rule.
Before using an AI application, you should carefully read its terms of use and privacy policy, for example, how the service uses the personal data you provide when you log in. Applications also save conversations you have with them, so think about what you say or write.
No sensitive or confidential information, such as personal data or trade secrets, should be given to AI applications. Any text you provide to the application will be used to improve its features. In other words, information may end up being used by outsiders.
The use of generative AI has significant environmental impacts because data centers consume large amounts of water and electricity.
A conversational, generative AI tool that answers questions and creates texts such as instructions, translations, etc. The University of Vaasa, Vaasa University of Applied Sciences and Novia University of Applied Sciences have a contract with Microsoft Copilot. Use primarily AI applications that your university has a contract with. This ensures that the application is reliable, secure and compatible with the university's privacy policies.
Many scientific databases acquired by the library have AI-powered features that support information searching like various assistants and search functions. They usually make information retrieval more efficient, for example, by suggesting appropriate articles or search terms for your topic.
Open AI's ChatGPT and Google's Gemini are also AI applications that generate human-like conversational responses.
Students can use artificial intelligence to support writing or editing text, but they must report the use of AI according to the guidelines of their own university. It is recommended to check with the teacher responsible for the course to see if AI applications can be used. The contents of theses must not be produced directly with AI, but the student must produce their own text. Check the instructions for your own university below.
