Content begins here
Main page content
Click to collapse
Deepening knowledge
Click to collapse
In this chapter, we’ll focus on one of these key skills: the effective use of generative AI and AI-powered tools. This technology opens up a wide range of opportunities for students to enhance their learning and explore innovative ways to gather information and solve problems. However, this is not a one-size-fits-all solution – so if you have any questions, always check with your academic program!
Legal Framework (EU AI Act) Click to collapse
EU AI Act
The EU AI Act is the first comprehensive legal framework governing the use of artificial intelligence in the European Union, which entered into force on August 1, 2024. The goal is to ensure that AI systems developed or used in the EU are safe, transparent, and compliant with fundamental rights. A “risk-based” approach is followed, meaning that the higher the potential risk of an AI system is classified, the stricter the requirements and obligations become. It is therefore a groundbreaking regulatory framework.
What this means for you and your studies is summarized below. You can find the detailed regulations at: EU AI Act
What are the different risk categories, and what do they mean?
According to the EU AI Act, AI systems are classified into four different risk categories.
Basics of AI Use
Artificial Intelligence is increasingly becoming part of everyday student life, for example in the form of text generators, translation tools, or assistive systems. The EU AI Act establishes a regulatory framework for this.
It pursues three central goals:
- Ensuring the responsible use of AI
- Protecting fundamental rights, particularly against discrimination
- Promoting transparency and traceability
In principle, this means that the use of AI is permitted – but the specific institutional regulations and guidelines of each organization always apply!
What does this mean for me as a student?
In addition to subject-specific knowledge, cross-disciplinary skills in dealing with Artificial Intelligence are becoming increasingly important. This includes, in particular, the reflective, responsible, and informed use of AI in your studies.
Specifically, this means:
- You have a basic understanding of how AI works and can generally comprehend and question the decisions made by AI systems.
- You are able to critically evaluate AI-generated content and do not rely exclusively on automated results, but maintain human oversight.
- You use AI tools in a targeted and thoughtful manner and document their use transparently in accordance with scientific standards.
- You handle data sensitively and help prevent discrimination through the responsible use of AI.
Opportunities and Risks of Generative AI Click to collapse
Opportunities and Risks of Generative AI
The EU AI Act regulates the responsible use of AI in research, business, and society, as well as digital security. This has direct and indirect implications for education and training at European universities and colleges. It ensures a values-based approach, so that AI systems are used in accordance with European fundamental values such as transparency, fairness, and accountability.
Against this backdrop, there are both opportunities and risks associated with the use of generative AI in education:
| Opportunities | Risks |
|---|---|
| Personalization of learning processes: AI can tailor learning content to individual prior knowledge, learning pace, and needs, thereby promoting self-directed learning. | Overreliance: Overreliance on AI can undermine critical thinking and human decision-making. |
| Support for cognitive processes: Students can use AI as a writing, learning, or reflection partner to better understand complex content and further develop their own ideas. | Hallucinations: AI models can generate false information that appears realistic but is not accurate. |
| Increased efficiency: Routine tasks such as summarizing, structuring, or drafting can be supported, leaving more time for in-depth learning. | Bias: Biased or insufficient training data leads to discriminatory results that disadvantage certain groups. |
| Encouraging of creativity and collaboration: AI can be used to spark ideas in brainstorming sessions or as a discussion partner, thereby opening up new perspectives. | Data privacy and copyright: Handling sensitive data can pose privacy risks if AI is not implemented correctly. |
For further information please refer to the recommendations of the Quick Start Recommendations of UNESCO IESALC (Institute for Higher Education in Latin America and the Caribbean)!
Your studies & AI Click to collapse
Your studies & AI
Proper prompting is essential for the effective use of generative AI. A prompt refers to a text input that is then processed by a machine – such as a computer or AI – to generate text, voice output, or images. A good prompt always depends on how it is phrased.
How to write a good prompt
To get the most accurate and effective response from generative AI, follow the “8 Ingredients for Successful Prompting.”
| 1. Persona | Assigning a specific role influences the nature, scope, tone, and complexity of the responses. Example: “You are a friendly, professional, and patient teaching assistant for a university statistics course. Your job is to help students with mechanics problems. Your answers should be brief and include specific suggestions for solving the problem.” |
| 2. Priming | Contextual information at the start of a chat, such as the problem statement, the AI’s task, objectives, response behavior, level, learning goals, etc. Example: “This concerns a statistics lecture at a university. The students have a basic understanding of statistics, and the task is at a high level of difficulty for them. The students are expected to develop an understanding of the solution steps...” |
| 3. Structural requirements | Guidelines regarding formal, structural, and stylistic aspects, such as instructions on writing style, subject-matter context, how to address users, formatting, etc. Example: “Write your answers in formal, academic language. Keep in mind that the users are from the field of economics...” |
| 4. Examples |
By providing examples, the AI can better understand the format in which the answers should be presented. Input: Black Now give me the answer for the following input: “Dark” |
| 5. Limit length | Specific instructions on how long the answer should be. Example: “Write a summary of text XY in 10 sentences for economics students aged 23–26...” |
| 6. Be precise / Avoid ambiguity | Avoid vague information (write in an academic style, but keep it accessible to young people) Example: “Write in an academic style, but use no more than two foreign words.” |
| 7. Markdown Syntax | Structure your prompt using Markdown syntax (sections marked with #) Example: # Your role: ..., # Your task: ..., # Instructions: .... |
| 8. Combination and Iteration | An iterative approach allows you to continually refine and clarify the prompt in order to find the most effective structure. A combination of different prompting strategies usually yields the best results. |
Source: Persike, M. (2023). Prompt-Labor: Generative KI in der Hochschullehre – Materialsammlung. Modul 1: Planungsphase. KI-Campus.
How can I use generative AI effectively?
Large language models, such as ChatGPT, have so far been viewed with some controversy. Some see great potential – when used correctly – while others harbor doubts about the credibility and usefulness of generative AI. What is clear is that AI can do more than simply summarize texts and generate “academic texts.” Current research shows that large language models offer a wide range of applications in education, but also present challenges.
Practical Application of Generative AI
Current discussions show that these applications must be integrated into educational contexts to avoid risks such as superficial learning or the uncritical acceptance of AI-generated content.
AI can simplify a wide range of tasks for us, including
- it can provide personalized learning support,
- promote collaboration and knowledge sharing,
- assist with time and organizational management, or
- support work on (scientific) texts.
If we view AI as a tool rather than a substitute for our own learning and thinking, it can help strengthen a variety of skills.
Here are a few examples of how you can use generative AI effectively for learning:
| Task | Format | Sample |
|---|---|---|
| AI as Learning Aid |
|
Creating a Study Plan:
Get an overview of the material: |
| Time and Organizational Management |
|
Structuring and Prioritizing: Here is my list and my current situation: [insert] Please provide me with: Important: |
| Collaboration and Knowledge Sharing |
|
Notes and Prioritization: You are my assistant for taking meeting notes and facilitating knowledge sharing. I will provide you with unorganized notes or bullet points from a conversation or meeting. Please organize them clearly, summarize the most important points concisely, and highlight key insights, decisions, and open issues. Formulate everything in a clear and concise manner without losing any important information. If appropriate, add a clear to-do list with responsibilities and next steps. Make sure the notes are easy to follow and suitable for sharing with the team. |
| Working on (academic) texts |
|
Text Optimization: 1. Correct grammar and spelling My text: Please provide me with: Important: |
Sources and further litertature
Free learning resources on artificial intelligence: https://www.ki-campus.org/. This site is in German you can take a few courses on English.
Land Nordrhein Westfalen (online). DigitalCheckNWR . Digital weiterwissen . https://www.digitalcheck.nrw/, retrieved on 23.01.2025.
Mollick , E., Mollick , L. (2023, 23. September). Assigning AI: Seven Approaches for Students, with Prompts . accessible via SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4475995 , retrieved on 08.04.2024.
Online Courses "Elements of AI": https://course.elementsofai.com/
Persike , M. (2023). Prompt Labor: Generative KI in der Hochschullehre Materialsammlung. Modul 1: Planungsphase. KI Campus.
UNESCO (2024). AI competency framework for students: https://unesdoc.unesco.org/ark:/48223/pf0000391105
https://youtu.be/VR9X9kRdgbk?si=3W4PmDvBBhC2UNN8
Learning Goals and AI
Throughout your studies, you will encounter what are known as learning objectives in all your courses. They help you measure your learning progress and assess your own learning. This means you can use the learning objectives to determine whether your current level of knowledge has already met or not met that objective. Learning objectives are usually formulated according to Bloom (1956).
Learning objectives thus make visible what you are actually supposed to be able to do in the learning process—and, above all, how you get there. They help you break down complex tasks into meaningful intermediate steps and better track your progress. When writing an academic text, for example, this means: You research literature, compare findings, critically evaluate them, and connect them to your research question. It is precisely in these steps that the actual learning takes place. With AI, many of these intermediate steps can be “outsourced”—such as summaries or initial arguments. But if you don’t have a clear focus on your learning objective, you’ll quickly end up delegating exactly the parts you should actually be learning yourself. Because the focus isn’t on the finished result, but on the path to getting there.
As a result, it is becoming increasingly important to formulate AI-related learning objectives as well. These objectives relate not only to subject-matter content but also to the competent use of AI tools. This includes, for example, critically evaluating results, identifying appropriate applications, and consciously designing one’s own work processes. Such learning goals help you use AI not just as a shortcut, but as a tool that specifically supports your learning process without replacing it.
AI competencies can also be analyzed based on this taxonomy of learning objectives. To do so, it is necessary to take into account the four pillars of AI competencies as defined by the EU AI Act.
- Basic AI Knowledge: Understanding that AI is designed by humans and that responsibility lies with humans.
- AI Selection & Evaluation: Selecting appropriate tools based on one’s own requirements and critically reflecting on the results.
- AI Ethics & Data Protection: Awareness of ethical issues, data protection, responsibility, and societal impacts.
- AI Application: Technical understanding of the use of AI tools in learning and everyday contexts.
Note: Bloom’s taxonomy originally comprises six levels (knowledge, comprehension, application, analysis, synthesis, evaluation). For the classification of AI learning objectives, this model is simplified and condensed into three levels: understanding, application, and creation. This simplification is made in this presentation because the specific classification ultimately depends on the specific wording of the learning objective.
| AI Comptence | Understanding | Application | Design |
|---|---|---|---|
| Basic Knowledge of AI | Foundational understanding of data, algorithms, and how AI applications work | Recognizing and assuming responsibility for AI applications; applying AI concepts to specific tasks | Taking a critical perspective on the (societal) impacts of AI use |
| AI Selection & Evaluation | Understand that different AI tools have different purposes, strengths, and limitations | Select appropriate AI tools and critically evaluate results | Critically compare AI systems and define use cases |
| AI Ethics & Data Protection | Basic understanding of ethics, bias, data protection, and social justice | Responsible use of AI, adherence to data protection and transparency | Actively integrate ethical principles into the use or design of AI |
| AI Applications | How AI can be used in educational and everyday contexts; when the use of AI should be avoided | Practical application of AI tools and evaluation of results | Use AI in a targeted and tailored manner for the task at hand, or adapt tools |