Artificial Intelligence in Teaching

Artificial Intelligence (AI) also influences digital teaching at the university. Here you will find an introduction to the topic.

What is AI?

Artificial intelligence (AI) is the ability of a machine to mimic human abilities such as reasoning, learning, planning and creativity. This brings challenges and potential. Here you will find a service offering with information about AI.

What is artificial intelligence?

Here is a compilation of interesting links on AI:

There are several projects that are trying to implement AI at universities. Here are some current examples:

  • Fobizz KI Tool
    Fobizz is a paid training platform for teachers in the school sector. It offers a very good paid interface to various generative AI generators. A prompt generator is also integrated, which can be used to easily create prompts for various teaching scenarios.
  • HAWKI Plattform
    The University of Applied Sciences and Arts in Hildesheim, Holzminden and Göttingen (HAWK) is developing the HAWKI platform as part of the Interaction Design degree programme. HAWKI acts as an interface for all university members to generative AI such as ChatGPT or Stable Diffusion. The prototype is based on ChatGPT 3 and is to be made available as open source on GitHub. HAWK uses a fee-based central interface (API) from OpenAI. The data sent via this interface is not analysed by LLM and is deleted after 30 days. The Hochschulforum Digitalisierung presents the project in an article.
  •  The University of Applied Sciences and Arts in Hildesheim, Holzminden and Göttingen (HAWK) is developing the HAWKI platform as part of the Interaction Design degree programme. HAWKI acts as an interface for all university members to generative AI such as ChatGPT or Stable Diffusion. The prototype is based on ChatGPT 3 and is to be made available as open source on GitHub. HAWK uses a fee-based central interface (API) from OpenAI. The data sent via this interface is not analysed by LLM and is deleted after 30 days. The Hochschulforum Digitalisierung presents the project in an article.
  • Ruhr University Bochum is currently trialling the implementation of open-source AI based on Llama 2 in Moodle courses.
  • RWTH Aachen University is trialling the implementation of an open-source AI in a Moodle research instance with the block plugin OpenAI Chat in cooperation with Azure Open AI and Microsoft.

General platforms and collections on AI

In this section you will find materials that you can use free of charge in your teaching, such as comprehensive self-study courses on AI, videos and other tips. You will also find tips on didactic application scenarios and on dealing with AI in university teaching.

  • Prompt-Labor Hochschullehre 2.0  by Malte Persike and Stefan Göllner, KI-Campus and Hochschulforum Digitalisierung, free self-study course on the following topics: Basics of generative AI, Introduction to effective prompting and prompting strategies, Quality assurance with GenAI, Personal assistants (custom GPTs), Prompting for teaching, Prompting for exams.
  • "The ChatGPT Prompt Book"
    Alan D. Thompson describes how to talk to ChatGPT so that the tool understands what we want in order to get better results. Learn how to get better at prompting and learn about basic and advanced prompts.
  • Learnprompting.org
    Free online courses to learn prompting in english, open source material.

 

"AI in University Teaching" ("KI in der Hochschullehre") is a video series on YouTube by the German association for university didactics (dghd). The videos are by and with renowned researchers and practitioners from higher education on the topic of AI.

  • How do I tell my AI? Background and principles of prompting at ChatGPT
    Teaser text: ChatGPT is an impressive artificial intelligence tool that allows you to communicate in natural language. Depending on how you talk to ChatGPT, you can get complex, substantive or creative responses. This talk will explain the importance of prompting in ChatGPT to guide dialogue with AI. Prompting is an important area also known as prompt competence, prompt engineering or prompt design. Although ChatGPT is currently much discussed, there are many misconceptions and myths that lead to ChatGPT being underestimated. In the lecture, such myths will first be cleared up in order to develop an understanding of "good prompting" and thus enable an optimal use of ChatGPT.
    Speaker: Prof. Dr. Michael Kipp (HS Augsburg)
    Length: 59:21
  • AI is more than text- and image-generating chatbots
    Teaser text: What types of AI are there and how can they be used on a micro, meso and macro level in higher education? In addition to input and discussion, there will be a collaborative part in which a "map" of AI projects in higher education institutions will be collected and made visible.
    Speaker: Prof. Dr. Niels Pinkwart (Humboldt University Berlin), Vice President for Teaching and Learning, Head of the Educational Technology Lab research area at the DFKI in Berlin.
    Length: 55:03
  • Text-generating tools such as ChatGPT make work easier for teachers
    Teaser text: Text-generating tools such as ChatGPT as work relief for teachers - how to use AI-supported tools to prepare your teaching.
    Speaker: PD Dr. Ulrike Hanke
    Length: 1:27:16
  • ChatGPT as a cognitive tool
    Teaser text: This interactive workshop is all about how ChatGPT can be used as a cognitive tool. The focus is on constructive and proactive ideas on how tools like ChatGPT can be used in higher education.
    Speaker: Prof. Dr. Christian Spannagel is professor of mathematics and its didactics with a focus on computer science and the implementation of new media at the Heidelberg University of Education.
    Length: 33:10
  • What are we here for? A value-based reflection and discussion on ChatGPT in research and teaching
    Teaser text: New technologies have always created challenges in teaching. The question is how to constructively incorporate tools like ChatGPT into teaching and use them responsibly. How to pass this on to the students and more will be discussed in this lecture.
    Speaker: Prof. Dr. Gabi Reinmann, Professor for Teaching and Learning at Universities at the University of Hamburg. In her research, she focuses on questions of higher education didactics in relation to new media or design-based research.
    Length: 53:50
  • ChatGPT for Non-Computer Scientists
    Teaser text: ChatGPT for non-computer scientists: Keys to understanding artificial intelligence and its applications in higher education.
    Speaker: Dr. Thomas Arnold, research associate in the field of "Ubiquitous Knowledge Processing" at the Department of Computer Science at Darmstadt University of Technology. In his research, he deals with the possibilities of current language models, as well as with the use of new language and text tools in university teaching.
    Length: 1:27:00
  • What is ChatGPT and how does it work? - And what similar tools are there?
    Teaser text: What is ChatGPT and how does it work? - And what similar tools are there?
    Speaker: Prof. Dr. Doris Weßels, Professor of Business Informatics at Kiel University of Applied Sciences, and co-founder of the Competence Centre "Teaching and Learning Writing with Artificial Intelligence - Tools and Techniques for Education and Science".
    Length: 1:22:59

Multimedia Kontor Hamburg makes training recordings available in thematic bundles:

Bildgenerator

  • Let’s chat & create! 30-minütige Einführung in ChatGPT und Dall-E
    Teaser text: In this 30-minute online training course, we introduce ChatGPT and DALL-E from Open AI. We look at the basic functions and possible applications of these two generative artificial intelligences, which can generate texts and images, among other things, but also talk about the challenges of using generative AI in higher education. No prior knowledge is required to participate in this course.
    Length: 32:49
  • Generative Fill! 30 Minuten zu: Bildbearbeitung in Photoshop mit Hilfe von KI
    Teaser text: The manipulation of images has long been possible with Photoshop and is becoming ever easier and more convincing. The next step is now to add or remove objects without tedious manual work, simply by entering text. Place a flock of birds in the sky? Remove people in the background? Everything is possible. What should your picture look like? Decide for yourself. This online course is aimed at beginners and no previous knowledge is required.
    Length: 23:56
  • It’s Magic! 30-minütige Einführung in Canva AI
    Teaser text: Introduction to the numerous AI functions of the Canva design tool. Canva is a comprehensive design tool that also offers numerous AI functions. In this 30-minute online training course, we will take a look at Magic Write, Magic Design, Magic Edit and Text-to-Image, among others. No prior knowledge is required to participate in this training.
    Length: 25:50
  • Let's be creative! 30-minütige Einführung in KI-Text-zu-Bild-Generatoren
    Teasertext: In dieser Online-Schulung geben wir einen kurzen Überblick über die bekanntesten Text-zu-Bild KI-Generatoren. Wir schauen uns deren Einsatzmöglichkeiten im Hochschulbereich an, einschließlich der Herausforderungen, die diese mit sich bringen können. Für die Teilnahme an dieser Schulung werden keine Vorkenntnisse vorausgesetzt.
    Länge: 22:35
  • Let’s create! 30-minütige Einführung in das AI-Text-zu-Bild-Tool Midjourney
    Teaser text: In this online training, we give a brief overview of the most popular text-to-image AI generators. We look at their potential uses in higher education, including the challenges they can present. No prior knowledge is required to attend this training.
    Length: 22:35
  • Text to Image! 30-minütige Einführung in die bildgebende KI „Adobe Firefly“
    Teaser text: Create unique images with everyday language by entering text. The web version of Adobe Firefly makes this very easy. Whether photorealistic or picturesque, Adobe Firefly offers many possibilities to generate unique images in a short time. Thanks to numerous setting options, the results can be refined further and further or quickly changed completely. In 30 minutes, we will get to know an exemplary workflow for image generation, which will guide us through the various work steps. At the end, we can switch to other Adobe applications with a generated image and continue working there.this online training is aimed at beginners and no previous knowledge is required.
    Length: 28:47
  • Unleash your Creativity! 30-minütige Einführung in das AI-Text-zu-Bild-Tool: Leonardo.AI 
    Teaser text: After we have already introduced Midjourney and DALL-E, in this 30-minute online training course we will talk about another text-to-image AI application: Leonardo.AI. We will show the interface, give an insight into prompting (text prompting) and how we can use it to create good images, as well as the possible applications in a university context. No prior knowledge is required to participate in this course.
    Length: 29:02

Videogenerator

  • Videoproduktion mit KI-Generatoren
    Teaser text: There is an ever-increasing demand for video content and video formats for a wide variety of distribution channels. Image optimization, post-production, text-to-video, video transcription, translation and subtitling are some of the key requirements for production and post-production. However, time-consuming and cost-intensive physical filming and days of post-production will be less and less necessary in order to create videos with high scalability. Is it already possible to create videos without professional knowledge, simply by entering text on AI generators and platforms? How can future video production processes be scaled and optimized?
    To what extent does it make sense to automate video production processes? And what is the potential for efficiency in the use of AI generators in terms of production costs, time and complexity? 
    This online training course presents AI generators for optimizing video production.
    Length: 51:47
  • Kreatives Storytelling mit KI! 30-minütige Einführung zu Deepfakes
    Teaser text: We create AI avatars and bring historical figures to life and tell stories. We use the AI applications ChatGPT, Midjourney, D-ID and ElevenLabs. No prior knowledge is required to participate in this course.
    Length: 22:53
  • Kreatives Storytelling mit KI! 30-minütige Einführung zu KI-Avataren
    Teaser text: We create AI avatars and bring historical figures to life and tell stories. We use the AI applications ChatGPT, Midjourney, D-ID and ElevenLabs. No prior knowledge is required to participate in this course.
    Length: 19:20
  • Let's generate Videos! 30 Min zu: Generative KI in der Videoproduktion
    Teaser text: What does the future of video production look like? This online training course gives you an overview of the latest applications of artificial intelligence in video production. It will show how AI tools are changing the way videos are created, edited and optimized. It is now even possible to use AI tools to generate short moving image sequences from images and text-to-video prompts. What could a future production workflow look like with the help of AI tools? This training course uses examples to show how generative AI is changing video production.
    Length: 33:26
  • Move it! 30-Minuten zu: Videogenerierung mit dem KI-Tool RunwayML
    Teaser text: Video clips can be generated from simple text or image input using the browser-based AI tool RunwayML and post-processed with Magic Tools. In 30 minutes you will learn an exemplary workflow for generating video clips from individual images or text prompts. You will be shown the individual work steps in RunwayML. This online training course is aimed at beginners and no prior knowledge is required.
    Length: 32:04
  • Next Level Post-Produktion! 30-Minuten zu: KI-Tools in DaVinci Resolve
    Teaser text: An overview of the current state of the art in the use of DaVinci Resolve AI tools.
    There are various AI tools in DaVinci Resolve that are designed to make post-production easier. In 30 minutes you will get an overview of the current state of the possibilities that already exist and that may soon be added. We will take a look at how the tools work and how they can be integrated into your own workflow. Basic knowledge of DaVinci Blackmagic Resolve is not required, but helpful.
    Length: 29:42

The teaser texts are taken from the videos of Multimedia Kontor Hamburg.

  • AI Campus
    A learning platform on artificial intelligence with online courses, videos and podcasts on the topic of AI. The main topics are chatbots & language assistants, data literacy, AI in medicine and AI in schools.
    Target group: Teachers, students and all interested parties.

 

  • TU Berlin
    Redesigning university teaching with artificial intelligence
    The TU Berlin describes a way of dealing with artificial intelligence in university teaching and explains, among other things, the background, working methods and legal framework for working with AI.
  • University of Potsdam
    The University of Potsdam writes about ChatGPT and Co. in university teaching. They explain what ChatGPT is and what it can do. They also give suggestions on how to deal with ChatGPT in teaching.

 

Prompt-Labor

Planning phase with AI

Here you can access the Moodle course of the Prompt Lab 
Module 1: Planning phase with AI

Kurs: Modul 1

Good practice at the FHP! AI in teaching

Here we present some "good practice" examples on the topic of AI in teaching, from committed teachers at the University of Applied Sciences Potsdam. They are sorted by date of publication. You will always find the most recent contributions at the top of the accordion.

If you would also like to publish an example of your teaching on the topic, please contact us at digitale-lehre@fh-potsdam.de.

In the winter semester 2023/2024, this cooperation course between HTW Berlin and Potsdam University of Applied Sciences, led by Prof Henrik Spohler and Prof Wiebke Loeper, explored the technologies currently being developed to produce photographs with the help of artificial intelligence.

What are the far-reaching possibilities and risks of artificial intelligence in the field of image production? Special attention was paid to the absurd basic grammar of AI-based image generation:

Every newly created image has its basis in the past because the technical intelligence has been trained with existing image material. We live in a time when we can no longer believe any digitally transmitted voice.

The focus of the project work is the new, difficult to grasp reference to reality of generated images. The result is over 25 different works about new worlds, lost stories, new traditions and the limits of artificial intelligence with critical observations.

 

A four-person team of researchers from various institutions analysed the use of ChatGPT by social work students. The Potsdam University of Applied Sciences was represented in the team by Dr Julius Späte and its students from various semesters of the social work degree programme contributed extensively to the data set.

In their study, the research team states that artificial intelligence systems are having a fundamental impact on the educational landscape and are expected to become increasingly relevant. The large language model ChatGPT in particular has led to a profound perturbation of the higher education landscape. ChatGPT has also found its way into social work studies and university members are faced with the challenge of developing an appropriate way of dealing with it. An essential starting point for developing a way of dealing with such systems can be found in a description of the current usage behaviour of students.

This quantitative study was dedicated to this task and offers such a starting point specifically for the field of social work studies. Across Germany and across all semesters, 875 social work students from various universities were surveyed on their usage behaviour of ChatGPT. In addition to usage behaviour, the study also surveyed students' evaluation of the AI tool, its thematisation in their studies and their affinity for technology. 
 

The results of the present research show a high usage of the tool at over 80% and thus emphasise the relevance of addressing the use of AI tools such as ChatGPT in social work studies and provide a basis for future research and didactic innovations.

Witter, S., Meinhardt-Injac, B., Siemer, L. & Späte, J. (2024). ChatGPT-Nutzung von Studierenden der Sozialen Arbeit. Eine quantitative Studie zur Nutzung, Bewertung und Thematisierung in der Hochschule aus Studierendensicht. 

 

In the course Design Against the Machine - Nonlinear Narratives of the Near Future by Prof Boris Müller, students are working intensively on interface design and AI.

In this course, students will work intensively with - and against - generative ‘AIs’. Experimental websites will be developed that play with the results of generative ‘AIs’ and provide a visionary outlook on the influences of ‘AIs’.

Artificial intelligence technologies have reached a new level, particularly in the area of text, code and image generation. Generative ‘AIs’ can produce results that are difficult to distinguish from man-made, creative works.

However, the use of ‘AI-generated’ products raises a number of legal, ethical and creative questions. It is obviously problematic to use texts and images that are based on the creative, intellectual and - above all - unpaid work of a large number of authors, illustrators or photographers. It is also problematic to use ‘AI-generated’ products and present them as the result of your own work. Generative ‘AIs’ call entire professions into question. Do we still need illustrators when Midjourney is cheaper and faster - and still creates high-quality images at the same time?

In this course, we want to take a targeted, proactive and creative look at the use of artificial intelligence in the design process. Based on concrete designs, we will reflect on legal, ethical and creative problems. We will use generative ‘AIs’ to create code, images and texts. These will be published in the form of an experimental website and then intensively processed, modified, destroyed and recombined as the course progresses.

Basic knowledge of HTML is required to participate in the course. However, workshops on HTML and CSS will also be offered during the course.

The seminar "AI in Film" in the winter semester 2203/24 for Bachelor students of European Media Studies (lecturers: Dr. Katrin von Kap-herr, FHP and Dr. Katharina Rein, UP) deals with the intertwining of media studies and AI in a multi-layered way. It promotes an open approach to AI through interactive discussions, critical reflection and the analysis of specialist literature and media artefacts such as films and series.

The focus is on critically analysing the concept and definitions of AI as well as ethical issues surrounding the use of AI. Practical applications and the use of AI in production are also analysed. The teaching concept contributes to the development of AI skills by enabling students to analyse complex issues, make informed decisions and critically reflect on the impact of AI on society.

About the project:

The aim of the research and development project "Any-Cubes" was to make it as easy as possible for children to play and experiment with artificial intelligence and networked objects. More information can be found on the project page Any-Cubes – Discovering Artificial Intelligence through Play.

Awards:

The project was awarded the delina-Award 2022 in the category "Early Childhood Education and School" at the LEARNTEC fair in Karlsruhe.

Press release on the delina Award for Any-Cubes

 

Lecturers from the Department of Information Sciences work together in small groups to test and reflect on the potential uses and challenges of generative AI for planning, implementing and evaluating their own teaching in a Prompt Lab, which was developed by the AI Campus as OER material. The Moodle course and the workbook of the Prompt Lab are adapted to specific subjects and needs in dialogue with the Dean's Office and can then be further developed independently by the Department of Information Sciences as OER material and format.

  • Module 1: Planning phase with AI – preparing lessons
  • Module 2: Implementation phase with AI – teaching scenarios
  • Module 3: Evaluation phase with AI – feedback and examinations

Trainer: Julia Lee

Format: Blended learning (approx. 1 hour online and 2 – 3 hours face-to-face per module)

Target group: Lecturers in the Department of Information Sciences at the FHP

Location and further information will be communicated via the Dean's Office of the Department of Information Sciences in the faculty.

This is a model project. If other departments are interested, a 30-minute presentation of the Prompt Lab concept can be requested; for the teaching conference of your own department via digitale-lehre@fh-potsdam.de.

The seminar "AI – but practical by Malte Völkner and Prof. Constanze Langer (Department of Design) dealt intensively with the use of AI in design processes in the summer semester 2023.

The students tested – also experimentally – the use and limits of current AI tools for various phases of a design process – from brainstorming and interface design to storytelling and logo design. They explored topics such as health, music, industrial design, law and education.

About the teaching concept:

The topic of artificial intelligence (AI) is a black box for many. The aim of the seminar was to unravel the secrets of the black box together, to find out what AI can do and to promote cooperation across disciplinary boundaries in classroom teaching.
The focus was always on social interaction at eye level, and classic dichotomies such as lecturer/student or expert/layperson were weakened or reversed.
Workshops with external guests on areas of expertise such as design, computer science, industry, art, ethics and journalism also enabled direct exchange with external AI practitioners to reflect on and understand basic concepts of current AI research (such as Deep Learning, Explainable AI) and to relate them to their own questions and specialist interests.
The learning objective was to develop technical, ethical and subject-specific basic knowledge of AI in order to identify points of contact for cooperative and interdisciplinary AI projects in their own field of work.

Awards:

In 2022, Dr. Katrin von Kap-herr and Dr. Alexander Scheidt received the state teaching award for their seminar "... but you are a robot, aren't you? Artificial Intelligence in Theory and Practice". This was an interdisciplinary Interflex seminar.

Press release on the state teaching award for Dr. Katrin von Kap-herr together with Dr. Alexander Scheidt

About the project:

The offer was developed by the Media Innovation Centre Babelsberg in cooperation with the seminar "AI Escape Room" by Dr Katrin von Kap-herr. The idea: to make the complex topic of AI, including its potential and challenges for the media world, accessible in a playful way. The project was awarded the audience award of the Potsdam prize for science communication in 2022 by proWissen Potsdam e.V..

The "AI.CUBE" is a live escape game with the aim of not only being a game about artificial intelligence (AI), but above all to provide an environment for knowledge transfer about AI and to explore which innovative possibilities for knowledge transfer can arise in an escape room. In workshops with experts and students of european media studies, indicators of gamification, playful media and storytelling were defined and topics related to AI were identified that can be translated into knowledge transfer elements via narrative and (audio-)visual tools. By playing the AI.CUBE - in contrast to commercial Escape Rooms - knowledge about artificial intelligence is conveyed to the players in a vivid and sustainable way.

Press release on the AI.CUBE Escape Room project

Awards:

The AI-Cube project won the 2022 audience award for science communication.

News about the Potsdam audience award for science communication for AI escape game

MaaS L.A.B.S. is a real-world project at the University of Applied Sciences Potsdam to develop a user-centred Mobility-as-a-Service platform with a project duration from may 2019 to May 2023.

The interdisciplinary MaaS L.A.B.S. project team from the social sciences, engineering and economics places transport users at the centre of its research and aims to test technologies and business models step by step in order to avoid planning errors. Four services are interlinked:

  • An app informs the citizens of a city about integrated mobility offers (public transport and sharing offers) and facilitates mobility planning and billing.
  • Intelligent sharing offers complement public transport to create an attractive overall offer that is competitive with private cars.
  • Demand-oriented shuttle buses pick up citizens at their doorsteps and thus improve the accessibility of large neighbourhoods.
  • Control systems ensure the smooth operation of this integrated transport system, which responds to current passenger requests in real time and in line with demand.

Together with the Berlin-based performance group Interrobang, the MaaS L.A.B.S. team invited people to a ride-sharing experiment on artificial intelligence and automated driving on 07 and 08 September 2021 in the north of Potsdam, simulating the use of an AI.

"Soulmachine – The self-learning Robobus" is a mobile interaction game between a group of passengers and an artificial intelligence. Five passengers take a seat in a converted minivan. Via tablet and voice control, they interact with an artificial intelligence and together take control of the vehicle. The machine develops the route and the driving experience from the collective decisions and behaviour of the passengers.

For more information, visit the Maas L.A.B.S. website.

Contact

Contact

Julia Lee, M. A. / Diplom
Studies and Teaching Unit
Head of the student e-learning team