6 Things to Consider Before Using Generative AI
Responsibly used, generative AI tools, such as Copilot, ChatGPT, Gemini or similar, can open up new opportunities for your learning and give you as a student support in everything from ideation and structure to understanding complex texts. In order to benefit from technology in a smart, safe, sustainable and permissible way, there are a few things that are important to know. Here we have gathered five key points that will help you get started with generative AI in your studies.
1. Use generative AI with a focus on learning
Strive to use generative AI tools in a way that empowers your own learning. Generative AI can be a good support to get started or move forward when you get stuck and need help to move forward. You can get help with structure, suggestions for content or feedback on your ideas. Generative AI can also answer questions, suggest arguments, proofread, compare texts, or summarize material to help you in a text or determine if it is relevant to your task. However, it is important to remember that generative AI should serve as a support in the work process and not as a substitute for your own work. Learning requires effort, and it is by processing the material yourself that you develop your understanding. Taking shortcuts by submitting texts that you have not written yourself rarely leads to any learning.
2. Always review information from generative AI
Texts generated by AI are not always factually accurate. Generative AI tools are designed to produce probable answers, not necessarily true ones, and can therefore provide misleading or inaccurate information, especially if they lack access to the right evidence. You should therefore always check the data against other reliable sources. By actively analyzing and reviewing AI-generated responses, you gradually develop a better understanding of the tools' possibilities and limitations.
3. Check if generative AI is allowed
If you plan to use generative AI in examinations, you always need to check with the responsible teacher whether it is allowed and how the use should be reported. For example, it may be about how and when generative AI tools should be listed in the list of sources. Even if it is not explicitly stated in the course information that generative AI is prohibited, it can still be considered cheating if you give the impression that something is your own work when in fact it has been generated by an AI tool.
4. Protect your privacy and that of others
To get good answers from generative AI, you need to formulate detailed and clear instructions and sometimes share texts or other material. It's important to be aware that what you input can be stored and used by the company behind the tool. Therefore, be careful about what you share and never enter information that is private, sensitive, or confidential. It may also be prohibited to upload copyrighted material, whether it is text, images, sound or video. Always check what is allowed and refrain from entering material if you are unsure.
5. Use generative AI sustainably
Generative AI tools require large amounts of energy and resources, both in development and in use. For example, using generative AI in a sustainable way can mean being aware of when and how you use the tools and whether they actually add any value to your learning. Therefore, avoid using generative AI casually or for things you might as well do in other ways, and instead try to formulate well-thought-out instructions that reduce the need for many repetitive tasks. By using generative AI in moderation and thoughtfulness, you not only contribute to your own learning, but also to more sustainable development.
6. Be vigilant of bias and cultural bias
Generative AI tools are trained on large amounts of data that can contain biases and cultural biases such as racism, hatred, misogyny and skewed descriptions of different groups. This means that the answers can sometimes reinforce stereotypes or provide a one-sided perspective. Therefore, try to critically review AI-generated texts and reflect on what norms, values, or assumptions may have affected the outcome.