Generative AI, Privacy and FOSS in Academic Research

I have been a university lecturer in Japan since 2009. I teach German and have mostly been researching comparative linguistics, readable fonts, and the like. However, my passion for privacy and cybersecurity has only grown stronger, and aside from my blogs and online courses, I really want to research and raise awareness about privacy on an academic level.

I created this picture using Create Studio 4.

This is, however, not an easy task. First, I am a language instructor, so it is kind of expected that I focus on language-related academic work. I am currently researching the use of LLMs in language learning, but since I am concerned about privacy, I wanted to give it an academic try in order to create ideal conditions and guidance for my students who want to use generative AI. It’s like preparing the soil before teaching someone how to grow crops.

I recently had the opportunity to present my research at the university in Japanese. I was even given a full hour for the presentation, followed by questions. I can say that things went well. I was able to justify my research. My colleagues asked many questions, but most of them were about the future of AI, superintelligence, and other potential threats. So they somewhat missed the point about privacy. Interestingly, however, they were very keen on FOSS solutions, which made me very happy, since I presented various FOSS solutions such as Brave Browser and Leo, as well as local AI models like Ollama. I also couldn’t resist pointing out that I was using a Linux machine during my presentation and that the slides were created with LibreOffice Impress. All in all, I am glad and thankful for the opportunity.

Here are the slides I used (in Japanese). Feel free to check them out.

I created the characters and the pictures with Create Studio 4 and the flipbook with Designrr.