Introduction
This semester, in the seminar “Fundamentals of Digital Communication,” our main focus was on artificial intelligence and its creative use. We engaged not only with the technical foundations and practice-oriented projects but also with the critical examination of the risks and opportunities of artificial intelligence.
Projects
Takeaways
Theoretical Foundations
This year as well, the topics followed one after another in quick succession. This led on the one hand to a kind of whiplash, but on the other hand also to an important gain in knowledge in the field of artificial intelligence.
At first, however, we dealt with the markup language Markdown, which makes it easy to structure texts clearly and neatly. Over the course of the semester, I realized, however, that I feel more comfortable working with HTML instead of Markdown. Even though Markdown is considerably easier and faster to use, I do not get into a proper workflow with it and feel somewhat more restricted.
The first real AI tool we got to know was Google’s Teachable Machine. This tool is very easy to use and thus a very beginner-friendly introduction to training AI models. It did not completely overwhelm us but clearly showed that AI is very dependent on the examples it is given. It also highlighted how important the quality and quantity of the input data are for the outcome.
Then we dealt with Large Language Models, a tool for generating texts. Particularly exciting for me in the technical examination was that LLMs at their core simply calculate the probability of the next word part (or token) and do not really think “intelligently.” So it is not really about the content of the texts that the tool writes, but only about the statistical relationships of the tokens.
After several weeks in which we dealt intensively with AI, the time also came when we examined the ethics behind the application of AI and weighed risks and opportunities in a debate.
In summary, one can say that the structure of the topics was very enriching. We touched on how AI actually works, how to properly prepare data for an AI, and also how the AI then processes this data. In addition, we also engaged with the ethical aspect.
Practical Application
This semester the theory was of course again connected with practical project work. In this context, we worked with image-generation, Teachable Machine, and LLMs, and integrated them into our websites in innovative ways. In addition, we dealt with the translation of natural patterns into code, which was not really related to the topic of AI but strongly challenged my creativity.
The projects as a whole all combined art and technology, and we were repeatedly reminded to create products that were as unusual as possible and not necessarily useful. I often had an idea for the project rather quickly and then only got a bit tangled up here and there during the implementation. But how so very often it was the case that one learns the most not from success but from the mistakes, which then have to be understood and fixed.
Every error in the code was always very exciting, and overall I had a lot of fun with practical coding. And because not too much was expected, we also had the time and calmness to experiment and explore what was actually possible.
Final
At the end of the semester, we once again had the task of compiling all projects into a web archive. I designed mine to be as clear and simple as possible. While most of the texts are in English, I deliberately chose to keep certain parts, such as the course material and the work process description for the Universal Interface project, in German. For me, it was important that the content provided lasting value, and it is simply easier in my native language to internalize concepts — especially when it comes to technical aspects.
Seeing all projects together on one website now gives me the sense that I have learned a great deal over the past months. Compared to the archive from the previous semester, I also feel that I have improved. At the same time, it motivates me to dive even deeper into the subject and to explore what more I can learn.